Search results for: teaching and learning empathy

Authors: Anudeep Appe, Bhanu Poluparthi, Lakshmi Kasivajjula, Udai Mv, Sobha Bagadi, Punya Modi, Aditya Singh, Hemanth Gunupudi, Spenser Troiano, Jeff Paul, Justin Stovall, Justin Yamamoto

Abstract:

The necessity of data-driven decisions in healthcare strategy formulation is rapidly increasing. A reliable framework which helps identify factors impacting a healthcare provider facility or a hospital (from here on termed as facility) market share is of key importance. This pilot study aims at developing a data-driven machine learning-regression framework which aids strategists in formulating key decisions to improve the facility’s market share which in turn impacts in improving the quality of healthcare services. The US (United States) healthcare business is chosen for the study, and the data spanning 60 key facilities in Washington State and about 3 years of historical data is considered. In the current analysis, market share is termed as the ratio of the facility’s encounters to the total encounters among the group of potential competitor facilities. The current study proposes a two-pronged approach of competitor identification and regression approach to evaluate and predict market share, respectively. Leveraged model agnostic technique, SHAP, to quantify the relative importance of features impacting the market share. Typical techniques in literature to quantify the degree of competitiveness among facilities use an empirical method to calculate a competitive factor to interpret the severity of competition. The proposed method identifies a pool of competitors, develops Directed Acyclic Graphs (DAGs) and feature level word vectors, and evaluates the key connected components at the facility level. This technique is robust since its data-driven, which minimizes the bias from empirical techniques. The DAGs factor in partial correlations at various segregations and key demographics of facilities along with a placeholder to factor in various business rules (for ex. quantifying the patient exchanges, provider references, and sister facilities). Identified are the multiple groups of competitors among facilities. Leveraging the competitors' identified developed and fine-tuned Random Forest Regression model to predict the market share. To identify key drivers of market share at an overall level, permutation feature importance of the attributes was calculated. For relative quantification of features at a facility level, incorporated SHAP (SHapley Additive exPlanations), a model agnostic explainer. This helped to identify and rank the attributes at each facility which impacts the market share. This approach proposes an amalgamation of the two popular and efficient modeling practices, viz., machine learning with graphs and tree-based regression techniques to reduce the bias. With these, we helped to drive strategic business decisions.

Keywords: competition, DAGs, facility, healthcare, machine learning, market share, random forest, SHAP

Procedia PDF Downloads 91

Authors: Katerina Florou

Abstract:

Two decades after coined the term "learner corpora" as collections of texts created by foreign or second language learners across various language contexts, and some years following suggestion to incorporate "focusing on form" within a Task-Based Learning framework, this study aims to explore how learner corpora, whether annotated with errors or not, can facilitate a focus on form in an educational setting. Argues that analyzing linguistic form serves the purpose of enabling students to delve into language and gain an understanding of different facets of the foreign language. This same objective is applicable when analyzing learner corpora marked with errors or in their raw state, but in this scenario, the emphasis lies on identifying incorrect forms. Teachers should aim to address errors or gaps in the students' second language knowledge while they engage in a task. Building on this recommendation, we compared the written output of two student groups: the first group (G1) employed the focusing on form phase by studying a specific aspect of the Italian language, namely the past participle, through examples from native speakers and grammar rules; the second group (G2) focused on form by scrutinizing their own errors and comparing them with analogous examples from a native speaker corpus. In order to test our hypothesis, we created four learner corpora. The initial two were generated during the task phase, with one representing each group of students, while the remaining two were produced as a follow-up activity at the end of the lesson. The results of the first comparison indicated that students' exposure to their own errors can enhance their grasp of a grammatical element. The study is in its second stage and more results are to be announced.

Keywords: Corpus interlanguage analysis, task based learning, Italian language as F1, learner corpora

Procedia PDF Downloads 54

Authors: Sushil Kumar Singh, Ashok Kumar, Ankit Ganeshpurkar, Ravi Singh, Devendra Kumar

Abstract:

In spectrum of neurodegenerative diseases, Alzheimer’s disease (AD) is characterized by the presence of amyloid β plaques and neurofibrillary tangles in the brain. It results in cognitive and memory impairment due to loss of cholinergic neurons, which is considered to be one of the contributing factors. Donepezil, an acetylcholinesterase (AChE) inhibitor which also inhibits butyrylcholinesterase (BuChE) and improves the memory and brain’s cognitive functions, is the most successful and prescribed drug to treat the symptoms of AD. The present work is based on designing of the selective BuChE inhibitors using computational techniques. In this work, machine learning models were trained using classification algorithms followed by screening of diverse chemical library of compounds. The various molecular modelling and simulation techniques were used to obtain the virtual hits. The amide derivatives of 4-(phenylsulfonamido) benzoic acid were synthesized and characterized using 1H & 13C NMR, FTIR and mass spectrometry. The enzyme inhibition assays were performed on equine plasma BuChE and electric eel’s AChE by method developed by Ellman et al. Compounds 31, 34, 37, 42, 49, 52 and 54 were found to be active against equine BuChE. N-(2-chlorophenyl)-4-(phenylsulfonamido)benzamide and N-(2-bromophenyl)-4-(phenylsulfonamido)benzamide (compounds 34 and 37) displayed IC50 of 61.32 ± 7.21 and 42.64 ± 2.17 nM against equine plasma BuChE. Ortho-substituted derivatives were more active against BuChE. Further, the ortho-halogen and ortho-alkyl substituted derivatives were found to be most active among all with minimal AChE inhibition. The compounds were selective toward BuChE.

Keywords: Alzheimer disease, butyrylcholinesterase, machine learning, sulfonamides

Procedia PDF Downloads 140

Authors: Miracle Ajah

Abstract:

One of the biggest challenges facing Theological/Religious Education in Nigeria is access to quality learning materials. For instance at the Trinity (Union) Theological College, Umuahia, it was difficult for lecturers to access suitable and qualitative materials for instruction especially the ones that would suit the African context and stimulate a deep rooted interest among the students. Some textbooks written by foreign authors were readily available in the School Library, but were lacking in the College bookshops for students to own copies. Even when the College was able to order some of the books from abroad, it did not usher in the needed enthusiasm expected from the students because they were either very expensive or very difficult to understand during private studies. So it became necessary to develop contextual materials which were affordable and understandable, though with little success. The National Open University of Nigeria (NOUN)’s innovation in the development and sharing of learning resources through its Open Course ware is a welcome development and of great assistance to students. Apart from NOUN students who could easily access the materials, many others from various theological/religious institutes across the nation have benefited immensely. So, the thesis of this paper is that the promotion of open educational resources in theological/religious education in Nigeria would facilitate a better informed/equipped religious leadership, which would in turn impact its adherents for a healthier society and national development. Adopting a narrative and historical approach within the context of Nigeria’s educational system, the paper discusses: educational traditions in Nigeria; challenges facing theological/religious education in Nigeria; and benefits of open educational resources. The study goes further to making recommendations on how OER could positively influence theological/religious education in Nigeria. It is expected that theologians, religious educators, and ODL practitioners would find this work very useful.

Keywords: OER, theological education, religious education, Nigeria

Procedia PDF Downloads 346

Authors: Shafait Hussain Ali

Abstract:

Since its outbreak in China, the Covid_19 19 disease has been caused by the corona virus SARS N coyote 2. Italy was the first Western country to be severely affected, and the first country to take drastic measures to control the disease. In start of December 2019, the sudden outbreaks of the Coronary Virus Disease was caused by a new Corona 2 virus (SARS-CO2) of acute respiratory syndrome in china city Wuhan. The World Health Organization declared the epidemic a public health emergency of international concern on January 30, 2020,. On February 14, 2020, 49,053 laboratory-confirmed deaths and 1481 deaths have been reported worldwide. The threat of the disease has forced most of the governments to implement various control measures. Therefore it becomes necessary to analyze the Italian data very carefully, in particular to investigates and to find out the present condition and the number of infected persons in the form of positive cases, death, hospitalized or some other features of infected persons will clear in simple form. So used such a model that will clearly shows the real facts and figures and also understandable to every readable person which can get some real benefit after reading it. The model used must includes(total positive cases, current positive cases, hospitalized patients, death, recovered peoples frequency rates ) all features that explains and clear the wide range facts in very simple form and helpful to administration of that country.

Keywords: machine learning tools and techniques, rapid miner tool, Naive-Bayes algorithm, predictions

Procedia PDF Downloads 108

Authors: Abdulmjeid Aljerawi

Abstract:

This study aimed to investigate the students' attitudes towards the use of social networking in education. Due to the nature of the study, and on the basis of its problem, objectives, and questions, the researcher used the descriptive approach. An appropriate questionnaire was prepared and validity and reliability were ensured. The questionnaire was then applied to the study sample of 434 students from King Saud University.

Keywords: social networks, education, learning, students

Procedia PDF Downloads 279

Authors: Klara Królewiak-Detsi, Anna Orylska, Anna Gorgolewska, Marta Boczkowska, Agata Graczykowska

Abstract:

This study investigates the communication between students and teachers in inclusive education in Poland. Specifically, we examine the communication and interaction of students with special educational needs during online learning compared to traditional face-to-face instruction. Our research questions are (1) how children with special educational needs communicate with their teachers and peers during online learning, and (2) what strategies can improve their communication skills. We conducted five focus groups with: (1) 55 children with special educational needs, (2) 65 typically developing pupils, (3) 28 professionals (psychologists and special education therapists), (4) 16 teachers, and (5) 16 parents of children with special educational needs. Our analysis focused on primary schools and used thematic analysis according to the 6-step procedure of Braun and Clarke. Our findings reveal that children with disabilities faced more difficulties communicating and interacting with others online than in face-to-face lessons. The online tools used for education were not adapted to the needs of children with disabilities, and schools lacked clear guidelines on how to pursue inclusive education online. Based on the results, we offer recommendations for online communication training and tools that are dedicated to children with special educational needs. Additionally, our results demonstrate that typically developing pupils are better in interpersonal relations and more often and effectively use social support. Children with special educational needs had similar emotional and communication challenges compared to their typically developing peers. In conclusion, our study highlights the importance of providing adequate support for the online education of children with special educational needs in inclusive classrooms.

Keywords: Inclusive education, Special educational needs, Social skills development, Online communication

Procedia PDF Downloads 132

Authors: Wei Deng

Abstract:

Based on sorting and analyzing related literature on CNKI for nearly three decades between 1987—2015, this paper, adopting the method of bibliometrics, summarized and reviewed the domestic research on sight translation from three aspects. The analysis concluded the following findings: 1) The majority research had focused on the noumenon of sight translation. The rest of the three main research perspectives are in descending order: sight translation teaching, sight translation skills and other associated skills, and cognitive research of sight translation. 2) The domestic research increased significantly in recent five years, but there is much room for the quality. 3) The non-empirical study has had higher proportion, while the empirical study is unitary with the lack of triangle validation. This paper suggested that sight translation being in sore need of unified definition, multilingual, even interdisciplinary cooperation.

Keywords: bibliometric analysis, perspectives, sight translation, tendency

Procedia PDF Downloads 336

Authors: Suphitcha Charoen-Amornkitt, Chokeanand Bussracumpakorn

Abstract:

The tendency of the Karen backstrap weaving succession has gradually decreased due to the difficulty of weaving techniques and the relocation of the young generation. The Yang Nam Klat Nuea community, Nong Ya Plong District, Phetchaburi, is a Pwo Karen community that is seriously confronted with a lack of cultural heritage. Thus, a group of weavers was formed to revive the knowledge of weaving. However, they have been gradually confronted with culture assimilation to mainstream culture from the desire for marketing acceptance and imperative and forced the extinction of culture due to the disappearance of weaving details and techniques. Although there are practical solutions, i.e., product development, community improvement, knowledge improvement, and knowledge transfer, to inherit the Karen weaving culture, people in the community cannot fulfill their deep intention about the weaving inheritance as most solutions have focused on developing the commercial products and making the income instead of inheriting their knowledge. This research employed qualitative user research with an in-depth user interview to study communal knowledge transfer succession based on the internal involved parties, i.e., four expert weavers, three young weavers, and three 4th generation villagers. The purpose is to explore the correlation and mindset of villagers towards the culture with specific issues, including the psychology of culture, core knowledge and learning methods, cultural inheritance, and cultural engagement. As a result, the existing models of knowledge management mostly focused on tangible strategies, which can notice progress in short terms, such as direct teaching and consistent practicing. At the same time, the motivation and passion of inheritors were abolished while the research found that the young generation who profoundly connected with the textile culture will have a more significant intention to continue the culture. Therefore, this research suggests both internal and external solutions to treat the community. Regarding the internal solutions, family, weaving group, and school have an important role to participate with young villagers by encouraging activities to support the cultivating of Karen’s history, understanding their identities, and adapting the culture as a part of daily life. At the same time, collecting all of the knowledge in the archives, e.g., recorded video, instruction, and books, can crucially prevent the culture from extinction. Regarding the external solutions, this study suggests that working with social media will enhance the intimacy of textile culture, while the community should relieve the roles in marketing competition and start to drive cultural experiences to create a new market position. In conclusion, this research intends to explore the causes and motivation to support the transfer of the culture to the 4th generation villagers and to raise awareness of the diversity of culture in society. With these suggestions and the desire to improve pride and confidence in culture, the community agrees that strengthening the relationships between the young villagers and the weaving culture can bring attention and interest back to the weaving culture.

Keywords: Pwo Karen textile culture, backstrap weaving succession, cultural inheritance, knowledge transfer, knowledge management

Procedia PDF Downloads 98

Authors: Michael Smalenberger

Abstract:

Intelligent tutoring systems (ITS) are computer-based platforms which can incorporate artificial intelligence to provide step-by-step guidance as students practice problem-solving skills. ITS can replicate and even exceed some benefits of one-on-one tutoring, foster transactivity in collaborative environments, and lead to substantial learning gains when used to supplement the instruction of a teacher or when used as the sole method of instruction. A common facet of many ITS is their use of Bayesian Knowledge Tracing (BKT) to estimate parameters necessary for the implementation of the artificial intelligence component, and for the probability of mastery of a knowledge component relevant to the ITS. While various techniques exist to estimate these parameters and probability of mastery, none directly and reliably ask the user to self-assess these. In this study, 111 undergraduate students used an ITS in a college-level introductory statistics course for which detailed transaction-level observations were recorded, and users were also routinely asked direct questions that would lead to such a self-assessment. Comparisons were made between these self-assessed values and those obtained using commonly used estimation techniques. Our findings show that such self-assessments are particularly relevant at the early stages of ITS usage while transaction level data are scant. Once a user’s transaction level data become available after sufficient ITS usage, these can replace the self-assessments in order to eliminate the identifiability problem in BKT. We discuss how these findings are relevant to the number of exercises necessary to lead to mastery of a knowledge component, the associated implications on learning curves, and its relevance to instruction time.

Keywords: Bayesian Knowledge Tracing, Intelligent Tutoring System, in vivo study, parameter estimation

Procedia PDF Downloads 174

Authors: Armin Nabaei, M. Omair Ahmad, M. N. S. Swamy

Abstract:

Many facial expression and emotion recognition methods in the traditional approaches of using LDA, PCA, and EBGM have been proposed. In recent years deep learning models have provided a unique platform addressing by automatically extracting the features for the detection of facial expression and emotions. However, deep networks require large training datasets to extract automatic features effectively. In this work, we propose an efficient emotion detection algorithm using face images when only small datasets are available for training. We design a deep network whose feature extraction capability is enhanced by utilizing several parallel modules between the input and output of the network, each focusing on the extraction of different types of coarse features with fined grained details to break the symmetry of produced information. In fact, we leverage long range dependencies, which is one of the main drawback of CNNs. We develop this work by introducing a Dynamic Soft-Margin SoftMax.The conventional SoftMax suffers from reaching to gold labels very soon, which take the model to over-fitting. Because it’s not able to determine adequately discriminant feature vectors for some variant class labels. We reduced the risk of over-fitting by using a dynamic shape of input tensor instead of static in SoftMax layer with specifying a desired Soft- Margin. In fact, it acts as a controller to how hard the model should work to push dissimilar embedding vectors apart. For the proposed Categorical Loss, by the objective of compacting the same class labels and separating different class labels in the normalized log domain.We select penalty for those predictions with high divergence from ground-truth labels.So, we shorten correct feature vectors and enlarge false prediction tensors, it means we assign more weights for those classes with conjunction to each other (namely, “hard labels to learn”). By doing this work, we constrain the model to generate more discriminate feature vectors for variant class labels. Finally, for the proposed optimizer, our focus is on solving weak convergence of Adam optimizer for a non-convex problem. Our noteworthy optimizer is working by an alternative updating gradient procedure with an exponential weighted moving average function for faster convergence and exploiting a weight decay method to help drastically reducing the learning rate near optima to reach the dominant local minimum. We demonstrate the superiority of our proposed work by surpassing the first rank of three widely used Facial Expression Recognition datasets with 93.30% on FER-2013, and 16% improvement compare to the first rank after 10 years, reaching to 90.73% on RAF-DB, and 100% k-fold average accuracy for CK+ dataset, and shown to provide a top performance to that provided by other networks, which require much larger training datasets.

Keywords: computer vision, facial expression recognition, machine learning, algorithms, depp learning, neural networks

Procedia PDF Downloads 75

Authors: Helen Durgante, Mariana F. Sparremberger, Flavia C. Bernardes, Debora D. DellAglio

Abstract:

Health promotion programmes for retirees, based on Positive Psychology perspectives for the development of strengths and virtues, demand broadened empirical investigation in Brazil. In the case of evidence-based applied research, it is suggested feasibility studies are conducted prior to efficacy trials of the intervention, in order to identify and rectify possible faults in the design and implementation of the intervention. The aim of this study was to evaluate the feasibility of a multicomponent Positive Psychology programme for health promotion of retirees, based on Cognitive Behavioural Therapy and Positive Psychology perspectives. The programme structure included six weekly group sessions (two hours each) encompassing strengths such as Values and self-care, Optimism, Empathy, Gratitude, Forgiveness, and Meaning of life and work. The feasibility criteria evaluated were: Demand, Acceptability, Satisfaction with the programme and with the moderator, Comprehension/Generalization of contents, Evaluation of the moderator (Social Skills and Integrity/Fidelity), Adherence, and programme implementation. Overall, 11 retirees (F=11), age range 54-75, from the metropolitan region of Porto Alegre-RS-Brazil took part in the study. The instruments used were: Qualitative Admission Questionnaire; Moderator Field Diary; the Programme Evaluation Form to assess participants satisfaction with the programme and with the moderator (a six-item 4-point likert scale), and Comprehension/Generalization of contents (a three-item 4-point likert scale); Observers’ Evaluation Form to assess the moderator Social Skills (a five-item 4-point likert scale), Integrity/Fidelity (a 10 item 4-point likert scale), and Adherence (a nine-item 5-point likert scale). Qualitative data were analyzed using content analysis. Descriptive statistics as well as Intraclass Correlations coefficients were used for quantitative data and inter-rater reliability analysis. The results revealed high demand (N = 55 interested people) and acceptability (n = 10 concluded the programme with overall 88.3% frequency rate), satisfaction with the program and with the moderator (X = 3.76, SD = .34), and participants self-report of Comprehension/Generalization of contents provided in the programme (X = 2.82, SD = .51). In terms of the moderator Social Skills (X = 3.93; SD = .40; ICC = .752 [IC = .429-.919]), Integrity/Fidelity (X = 3.93; SD = .31; ICC = .936 [IC = .854-.981]), and participants Adherence (X = 4.90; SD = .29; ICC = .906 [IC = .783-.969]), evaluated by two independent observers present in each session of the programme, descriptive and Intraclass Correlation results were considered adequate. Structural changes were introduced in the intervention design and implementation methods, as well as the removal of items from questionnaires and evaluation forms. The obtained results were satisfactory, allowing changes to be made for further efficacy trials of the programme. Results are discussed taking cultural and contextual demands in Brazil into account.

Keywords: feasibility study, health promotion, positive psychology intervention, programme evaluation, retirees

Procedia PDF Downloads 196

Authors: Yan Zhang

Abstract:

Predictive maintenance is a technique to predict when an in-service machine will fail so that maintenance can be planned in advance. Analytics-driven predictive maintenance is gaining increasing attention in many industries such as manufacturing, utilities, aerospace, etc., along with the emerging demand of Internet of Things (IoT) applications and the maturity of technologies that support Big Data storage and processing. This study aims to build an end-to-end analytics solution that includes both real-time machine condition monitoring and machine learning based predictive analytics capabilities. The goal is to showcase a general predictive maintenance solution architecture, which suggests how the data generated from field machines can be collected, transmitted, stored, and analyzed. We use a publicly available aircraft engine run-to-failure dataset to illustrate the streaming analytics component and the batch failure prediction component. We outline the contributions of this study from four aspects. First, we compare the predictive maintenance problems from the view of the traditional reliability centered maintenance field, and from the view of the IoT applications. When evolving to the IoT era, predictive maintenance has shifted its focus from ensuring reliable machine operations to improve production/maintenance efficiency via any maintenance related tasks. It covers a variety of topics, including but not limited to: failure prediction, fault forecasting, failure detection and diagnosis, and recommendation of maintenance actions after failure. Second, we review the state-of-art technologies that enable a machine/device to transmit data all the way through the Cloud for storage and advanced analytics. These technologies vary drastically mainly based on the power source and functionality of the devices. For example, a consumer machine such as an elevator uses completely different data transmission protocols comparing to the sensor units in an environmental sensor network. The former may transfer data into the Cloud via WiFi directly. The latter usually uses radio communication inherent the network, and the data is stored in a staging data node before it can be transmitted into the Cloud when necessary. Third, we illustrate show to formulate a machine learning problem to predict machine fault/failures. By showing a step-by-step process of data labeling, feature engineering, model construction and evaluation, we share following experiences: (1) what are the specific data quality issues that have crucial impact on predictive maintenance use cases; (2) how to train and evaluate a model when training data contains inter-dependent records. Four, we review the tools available to build such a data pipeline that digests the data and produce insights. We show the tools we use including data injection, streaming data processing, machine learning model training, and the tool that coordinates/schedules different jobs. In addition, we show the visualization tool that creates rich data visualizations for both real-time insights and prediction results. To conclude, there are two key takeaways from this study. (1) It summarizes the landscape and challenges of predictive maintenance applications. (2) It takes an example in aerospace with publicly available data to illustrate each component in the proposed data pipeline and showcases how the solution can be deployed as a live demo.

Keywords: Internet of Things, machine learning, predictive maintenance, streaming data

Procedia PDF Downloads 387

Authors: Tassadit Issaadi Hamitouche, Nicolas Gillard, Jean Petit, Valerie Lavaste, Celine Mayousse

Abstract:

Water is a fundamental resource for the industry. It is taken from the environment either from municipal distribution networks or from various natural water sources such as the sea, ocean, rivers, aquifers, etc. Once used, water is discharged into the environment, reprocessed at the plant or treatment plants. These withdrawals and discharges have a direct impact on natural water resources. These impacts can apply to the quantity of water available, the quality of the water used, or to impacts that are more complex to measure and less direct, such as the health of the population downstream from the watercourse, for example. Based on the analysis of data (meteorological, river characteristics, physicochemical substances), we wish to predict water stress episodes and anticipate prefectoral decrees, which can impact the performance of plants and propose improvement solutions, help industrialists in their choice of location for a new plant, visualize possible interactions between companies to optimize exchanges and encourage the pooling of water treatment solutions, and set up circular economies around the issue of water. The development of a system for the collection, processing, and use of data related to water resources requires the functional constraints specific to the latter to be made explicit. Thus the system will have to be able to store a large amount of data from sensors (which is the main type of data in plants and their environment). In addition, manufacturers need to have 'near-real-time' processing of information in order to be able to make the best decisions (to be rapidly notified of an event that would have a significant impact on water resources). Finally, the visualization of data must be adapted to its temporal and geographical dimensions. In this study, we set up an infrastructure centered on the TICK application stack (for Telegraf, InfluxDB, Chronograf, and Kapacitor), which is a set of loosely coupled but tightly integrated open source projects designed to manage huge amounts of time-stamped information. The software architecture is coupled with the cross-industry standard process for data mining (CRISP-DM) data mining methodology. The robust architecture and the methodology used have demonstrated their effectiveness on the study case of learning the level of a river with a 7-day horizon. The management of water and the activities within the plants -which depend on this resource- should be considerably improved thanks, on the one hand, to the learning that allows the anticipation of periods of water stress, and on the other hand, to the information system that is able to warn decision-makers with alerts created from the formalization of prefectoral decrees.

Keywords: data mining, industry, machine Learning, shortage, water resources

Procedia PDF Downloads 122

Authors: Meng Su

Abstract:

High-dimensional data analysis often presents challenges in capturing the complex, nonlinear relationships and manifold structures inherent to the data. This article presents a novel approach that leverages the strengths of two powerful techniques, Diffusion Maps and Diffusion Probabilistic Models (DPMs), to address these challenges. By integrating the dimensionality reduction capability of Diffusion Maps with the data modeling ability of DPMs, the proposed method aims to provide a comprehensive solution for analyzing and generating high-dimensional data. The Diffusion Map technique preserves the nonlinear relationships and manifold structure of the data by mapping it to a lower-dimensional space using the eigenvectors of the graph Laplacian matrix. Meanwhile, DPMs capture the dependencies within the data, enabling effective modeling and generation of new data points in the low-dimensional space. The generated data points can then be mapped back to the original high-dimensional space, ensuring consistency with the underlying manifold structure. Through a detailed example implementation, the article demonstrates the potential of the proposed hybrid approach to achieve more accurate and effective modeling and generation of complex, high-dimensional data. Furthermore, it discusses possible applications in various domains, such as image synthesis, time-series forecasting, and anomaly detection, and outlines future research directions for enhancing the scalability, performance, and integration with other machine learning techniques. By combining the strengths of Diffusion Maps and DPMs, this work paves the way for more advanced and robust data analysis methods.

Keywords: diffusion maps, diffusion probabilistic models (DPMs), manifold learning, high-dimensional data analysis

Procedia PDF Downloads 111

Authors: Nicole Merkle, Stefan Zander

Abstract:

Ambient Assisted Living (AAL) describes a technological and methodological stack of (e.g. formal model-theoretic semantics, rule-based reasoning and machine learning), different aspects regarding the behavior, activities and characteristics of humans. Hence, a semantic representation of the user environment and its relevant elements are required in order to allow assistive agents to recognize situations and deduce appropriate actions. Furthermore, the user and his/her characteristics (e.g. physical, cognitive, preferences) need to be represented with a high degree of expressiveness in order to allow software agents a precise evaluation of the users’ context models. The correct interpretation of these context models highly depends on temporal, spatial circumstances as well as individual user preferences. In most AAL approaches, model representations of real world situations represent the current state of a universe of discourse at a given point in time by neglecting transitions between a set of states. However, the AAL domain currently lacks sufficient approaches that contemplate on the dynamic adaptions of context-related representations. Semantic representations of relevant real-world excerpts (e.g. user activities) help cognitive, rule-based agents to reason and make decisions in order to help users in appropriate tasks and situations. Furthermore, rules and reasoning on semantic models are not sufficient for handling uncertainty and fuzzy situations. A certain situation can require different (re-)actions in order to achieve the best results with respect to the user and his/her needs. But what is the best result? To answer this question, we need to consider that every smart agent requires to achieve an objective, but this objective is mostly defined by domain experts who can also fail in their estimation of what is desired by the user and what not. Hence, a smart agent has to be able to learn from context history data and estimate or predict what is most likely in certain contexts. Furthermore, different agents with contrary objectives can cause collisions as their actions influence the user’s context and constituting conditions in unintended or uncontrolled ways. We present an approach for dynamically updating a semantic model with respect to the current user context that allows flexibility of the software agents and enhances their conformance in order to improve the user experience. The presented approach adapts rules by learning sensor evidence and user actions using probabilistic reasoning approaches, based on given expert knowledge. The semantic domain model consists basically of device-, service- and user profile representations. In this paper, we present how this semantic domain model can be used in order to compute the probability of matching rules and actions. We apply this probability estimation to compare the current domain model representation with the computed one in order to adapt the formal semantic representation. Our approach aims at minimizing the likelihood of unintended interferences in order to eliminate conflicts and unpredictable side-effects by updating pre-defined expert knowledge according to the most probable context representation. This enables agents to adapt to dynamic changes in the environment which enhances the provision of adequate assistance and affects positively the user satisfaction.

Keywords: ambient intelligence, machine learning, semantic web, software agents

Procedia PDF Downloads 282

Authors: Tianqi Wu, Min Wang

Abstract:

Research investigating language acquisition from a constructionist perspective has demonstrated that language is learned as constructions at various linguistic levels, which is related to factors of frequency, semantic prototypicality, and form-meaning contingency. However, previous research on construction learning tended to focus on clause-level constructions such as verb argument constructions but few attempts were made to study morpheme-level constructions such as the progressive construction, which is regarded as a source of acquisition problems for English learners from diverse L1 backgrounds, especially for those whose L1 do not have an equivalent construction such as German and Chinese. To trace the developmental trajectory of Chinese EFL learners’ use of the progressive with respect to verb frequency, verb-progressive contingency, and verbal prototypicality and generality, a learner corpus consisting of three sub-corpora representing three different English proficiency levels was extracted from the Chinese Learners of English Corpora (CLEC). As the reference point, a native speakers’ corpus extracted from the Louvain Corpus of Native English Essays was also established. All the texts were annotated with C7 tagset by part-of-speech tagging software. After annotation all valid progressive hits were retrieved with AntConc 3.4.3 followed by a manual check. Frequency-related data showed that from the lowest to the highest proficiency level, (1) the type token ratio increased steadily from 23.5% to 35.6%, getting closer to 36.4% in the native speakers’ corpus, indicating a wider use of verbs in the progressive; (2) the normalized entropy value rose from 0.776 to 0.876, working towards the target score of 0.886 in native speakers’ corpus, revealing that upper-intermediate learners exhibited a more even distribution and more productive use of verbs in the progressive; (3) activity verbs (i.e., verbs with prototypical progressive meanings like running and singing) dropped from 59% to 34% but non-prototypical verbs such as state verbs (e.g., being and living) and achievement verbs (e.g., dying and finishing) were increasingly used in the progressive. Apart from raw frequency analyses, collostructional analyses were conducted to quantify verb-progressive contingency and to determine what verbs were distinctively associated with the progressive construction. Results were in line with raw frequency findings, which showed that contingency between the progressive and non-prototypical verbs represented by light verbs (e.g., going, doing, making, and coming) increased as English proficiency proceeded. These findings altogether suggested that beginning Chinese EFL learners were less productive in using the progressive construction: they were constrained by a small set of verbs which had concrete and typical progressive meanings (e.g., the activity verbs). But with English proficiency increasing, their use of the progressive began to spread to marginal members such as the light verbs.

Keywords: Construction learning, Corpus-based, Progressives, Prototype

Procedia PDF Downloads 128

Authors: Eva Nwokah, Casey Taliancich-Klinger, Lauren Luna, Sarah Rodriguez

Abstract:

Story-retelling is a technique frequently used to assess children’s language skills and support their development of narratives. Fourteen preschool children listened to one of two stories from the wordless, illustrated Frog book series and then retold the story using the pictures. A comparison of three verb types (action, mental and other) in the original story model, and children's verb use in their retold stories revealed the salience of action events. The children's stories contained a similar proportion of verb types to the original story. However, the action verbs they used were rarely those they had heard in the original. The implications for the process of lexical encoding and narrative recall are discussed, as well as suggestions for the use of wordless picture books and the language teaching of new verbs.

Keywords: story re-telling, verb use, preschool language, wordless picture books

Procedia PDF Downloads 272

Authors: Bernard Igoche Igoche, Olumuyiwa Matthew, Peter Bednar, Alexander Gegov

Abstract:

This study presents an ontology-driven approach for knowledge discovery and validation from admission databases in Nigerian polytechnic institutions. The research aims to address the challenges of extracting meaningful insights from vast amounts of admission data and utilizing them for decision-making and process improvement. The proposed methodology combines the knowledge discovery in databases (KDD) process with a structural causal model (SCM) ontological framework. The admission database of Benue State Polytechnic Ugbokolo (Benpoly) is used as a case study. The KDD process is employed to mine and distill knowledge from the database, while the SCM ontology is designed to identify and validate the important features of the admission process. The SCM validation is performed using the conditional independence test (CIT) criteria, and an algorithm is developed to implement the validation process. The identified features are then used for machine learning (ML) modeling and prediction of admission status. The results demonstrate the adequacy of the SCM ontological framework in representing the admission process and the high predictive accuracies achieved by the ML models, with k-nearest neighbors (KNN) and support vector machine (SVM) achieving 92% accuracy. The study concludes that the proposed ontology-driven approach contributes to the advancement of educational data mining and provides a foundation for future research in this domain.

Keywords: admission databases, educational data mining, machine learning, ontology-driven knowledge discovery, polytechnic education, structural causal model

Procedia PDF Downloads 66

Authors: Noriyuki Suyama

Abstract:

The progress and spread of digitalization have led to the provision of a variety of new services. The recent progress in digitization can be attributed to rapid developments in science and technology. First, the research and diffusion of artificial intelligence (AI) has made dramatic progress. Around 2000, the third wave of AI research, which had been underway for about 50 years, arrived. Specifically, machine learning and deep learning were made possible in AI, and the ability of AI to acquire knowledge, define the knowledge, and update its own knowledge in a quantitative manner made the use of big data practical even for commercial PCs. On the other hand, with the spread of social media, information exchange has become more common in our daily lives, and the lending and borrowing of goods and services, in other words, the sharing economy, has become widespread. The scope of this trend is not limited to any industry, and its momentum is growing as the SDGs take root. In addition, the Social Network Service (SNS), a part of social media, has brought about the evolution of the retail business. In the past few years, social network services (SNS) involving users or companies have especially flourished. The People's Republic of China (hereinafter referred to as "China") is a country that is stimulating enormous consumption through its own unique SNS, which is different from the SNS used in developed countries around the world. This paper focuses on the effectiveness and challenges of influencer marketing by focusing on the influence of influencers on users' behavior and satisfaction with Chinese SNSs. Specifically, Conducted was the quantitative survey of Tik Tok users living in China, with the aim of gaining new insights from the analysis and discussions. As a result, we found several important findings and knowledge.

Keywords: customer satisfaction, social networking services, influencer marketing, Chinese consumers’ behavior

Procedia PDF Downloads 89

Authors: Mahmoud Nabilu

Abstract:

The present study explored the effect of using Mobile listening applications on developing listening skills by Iranian intermediate EFL learners. Fifty male intermediate English learners whose age range was between 15 and 20, participated in the study. The participants were placed in two groups on the basis of their scores on a placement test. Therefore, the participants of the study were homogenized in terms of general proficiency, and groups were assigned as one experimental group and one control group. The experimental group was instructed by the treatment which was using mobile applications to develop their listening skills while the control group received traditional methods. The research data were obtained from the 40-item multiple-choice tests as a pre-test and a post-test. The results of the t-test clearly revealed that the learners in the experimental group performed better in the post-test than the pre-test. This implies that using a mobile application for developing listening skills as a treatment was effective in helping the language learners perform better on post-test. However, a statistically significant difference was found between the post-tests scores of the two groups. The mean of the experimental group was greater compared to the control group. The participants were Iranian and from an Iranian Language Institute, so care should be taken while generalizing the results to the learners of other nationalities. However, in the researcher's view, the findings of this study have valuable implications for teachers and learners, methodologists and syllabus designers, linguists and MALL/CALL (mobile/computer-assisted language learning) experts. Using the result of the present paper is an aim of raising the consciousness of a better technique of developing listening skills in order to make language learning more efficient for the learners.

Keywords: Mobile listening applications, intermediate EFL learners, MALL, CALL

Procedia PDF Downloads 195

Authors: Indujeeva Keerthila Peiris

Abstract:

The secular view of mindfulness has some connotation to the original meaning of mindfulness mentioned in the Theravada Buddhist texts (Pāli Canon), but there is a substantial difference in the meaning of the two. Secular Mindfulness Based Interventions (MBI) focus on stilling the mind, which may provide short-term benefits and help individuals to deal with physical pain, grief, and distress. However, as with many popular educational innovations, the foundational values of mindfulness strategies have been distorted and subverted in a number of instances in which ‘McMindfulness’ programmes have been implemented with a view to reducing mindfulness mediation as a self-help technique that is easily misappropriated for the exclusive pursuit of corporate objectives, employee pacification, and commercial profit. The intention of this paper is not to critique the misappropriations of mindfulness. Instead, to go back to the root source and bring insights from the Buddhist Pāli Canon and its associated teachings on mindfulness in its own terms. In the Buddha’s discourses, as preserved in the Pāli Canon, there is nothing more significant than the understanding and practice of ‘Satipatthãna’. The Satipatthāna Sutta , the ‘Discourse on the Establishment of Mindfulness,’ opens with a proclamation highlighting both the purpose of this training and its methodology. The right practice of mindfulness is the gateway to understanding the Buddha’s teaching. However, although this concept is widely discussed among the Dhamma practitioners, it is the least understood one of them all. The purpose of this paper is to understand deeper meaning of mindfulness as it was originally intended by the Teacher. The natural state of mind is that it wanders. It wanders into the past, the present, and the future. One’s ability to hold attention to a mind object (emotion, thought, feeling, sensation, sense impression) called ‘concentration’. The intentional concentration process does not lead to wisdom. However, the development of wisdom starts when the mind is calm, concentrated, and unified. The practice of insight contemplation aims at gaining a direct understanding of the real nature of phenomena. According to the Buddha’s teaching, there are three basic facts of all existence: 1) impermanence (anicca in Pāli) ; 2) fabrication (also commonly known as suffering, unsatisfactoriness, sankhara or dukka in Pāli); 3) not-self (insubstantiality or impersonality, annatta in Pāli ). The entire Buddhist doctrine is based on these three facts. The problem is our ignorance covers reality. It is not that a person sees the emptiness of them or that we try to see the emptiness of our experience by conceptually thinking that they are empty. It is an experiential outcome that happens when the cause-and- effect overrides the self-view (sakkaya dhitti), and ignorance is known as ignorance and eradicated once and for all. Therefore, the right view (samma dhitti) is the starting point of the path, not ethical conduct (sila) or samadhi (jhana). In order to develop the right view, we need to first listen to the correct Dhamma and possess Yoniso manasikara (right comprehension) to know the five aggregates as five aggregates.

Keywords: mindfulness, spirituality, buddhism, pali canon

Procedia PDF Downloads 78

Authors: Carol S. Binnekade, Deon Scott, Christina C. Shuttleworth, Annelien A. Van Rooyen

Abstract:

Higher education institutions are constantly challenged to deliver skilled graduates into the workplace. Employers expect graduates to have the required technical knowledge as well as various pervasive skills. This also applies to accountants who need to know the technical requirements of financial reporting and be able to communicate with individuals, teams and clients at a high level. Accountants need to develop effective business conversational skills and use these skills to communicate up, down and across organizations, taking into consideration cultural and gender diversity. In addition, they need to master business writing and presentation skills. However, providing students with these skills in a distance-learning environment where interaction between students and instructors is limited, is a challenge for academics. The study on which this paper reports, forms part of a larger body of research, which explored the perceptions of accounting practitioners of the communication skills (or lack thereof) of recently qualified accounting students. Feedback (qualitative and quantitative) was obtained from various accounting practitioners in South Africa. Taking into consideration that distance learners communicate mainly with their instructors via email communication and their assignments are submitted using various word processor software, the researchers were of the opinion that the accounting graduates would be capable of communicating effectively once they entered the workplace. However, the research findings, inter alia, suggested that the accounting graduates lacked communication skills and that training was needed to differentiate between business and social communication once they entered the workplace. Recommendations on how these communication challenges may be addressed by higher education institutions are provided.

Keywords: accounting practitioners, communication skills, distance education, pervasive skills

Procedia PDF Downloads 204

Authors: Alvin Uchenna Ugwu

Abstract:

The United Nation’s 2030 agenda and Global Action Programme (GAP) for implementation of the Sustainable Development Goals (SDGs), has mandated all sectors in the societies, including education, to develop strategies towards actualizing sustainability in all facets of the society, by the year 2030. Education is no doubt a key tool for social change. However, educational institutions in most African nations need a paradigmatic shift to strike a balance between policies (curricular) and practices, with regards to Education for Sustainable Development (ESD). The paradigm shift in this regard is described as whole-institution/school approach. The whole institution approaches advocate action-focused ESD. In other words, ESD policy and curriculum makers, formal and non-formal education institutions, need to ‘practice what they preach’. This paper is developed from an ongoing study carried out by the author and guided by two research questions: -What are the views of intermediate phase science and technology preservice teachers on the ESD content included in the science and technology modules? -What challenges or enable intermediate phase science and technology pre-service teachers to learn about ESD in science and technology modules? The study drew from the views and experiences of preservice science teachers, learning about ESD in a university’s college of education in South Africa. Using qualitative case study research design, the research data were generated via questionnaires and focus group discussions. Analysis of generated data indicates that universities and institutions of higher learning need to demonstrate practical involvement while implementing ESD in societies, rather than just standing as knowledge media. Findings of the study further suggest that natural sciences and technology courses in teacher education programmes and other institutions of higher learning, should be perceived as key transformative tools in shaping the consciousness of students towards integrating and fostering ESD in developing countries such as South Africa. Thus, this paper seeks to promote ‘Whole Institution Involvement’ in teacher education colleges in South Africa, as a measure of improving ESD in higher education settings. The paper suggests that in order to achieve ESD in higher education settings and beyond, policies and practices should be reexamined beyond rhetoric and buzzwords. The paper further argues that implementation of ESD is largely influenced by context, hence two different contexts should be examined empirically.

Keywords: education for sustainable development, higher education institutions, pre-service science teachers, qualitative case study research, whole institution involvement

Procedia PDF Downloads 177

Authors: Jeremy Jenrette, Z. Y. C. Liu, Pranav Chimote, Edward Fox, Trevor Hastie, Francesco Ferretti

Abstract:

Suitable shark conservation depends on well-informed population assessments. Direct methods such as scientific surveys and fisheries monitoring are adequate for defining population statuses, but species-specific indices of abundance and distribution coming from these sources are rare for most shark species. We can rapidly fill these information gaps by boosting media-based remote monitoring efforts with machine learning and automation. We created a database of shark images by sourcing 24,546 images covering 219 species of sharks from the web application spark pulse and the social network Instagram. We used object detection to extract shark features and inflate this database to 53,345 images. We packaged object-detection and image classification models into a Shark Detector bundle. We developed the Shark Detector to recognize and classify sharks from videos and images using transfer learning and convolutional neural networks (CNNs). We applied these models to common data-generation approaches of sharks: boosting training datasets, processing baited remote camera footage and online videos, and data-mining Instagram. We examined the accuracy of each model and tested genus and species prediction correctness as a result of training data quantity. The Shark Detector located sharks in baited remote footage and YouTube videos with an average accuracy of 89\%, and classified located subjects to the species level with 69\% accuracy (n =\ eight species). The Shark Detector sorted heterogeneous datasets of images sourced from Instagram with 91\% accuracy and classified species with 70\% accuracy (n =\ 17 species). Data-mining Instagram can inflate training datasets and increase the Shark Detector’s accuracy as well as facilitate archiving of historical and novel shark observations. Base accuracy of genus prediction was 68\% across 25 genera. The average base accuracy of species prediction within each genus class was 85\%. The Shark Detector can classify 45 species. All data-generation methods were processed without manual interaction. As media-based remote monitoring strives to dominate methods for observing sharks in nature, we developed an open-source Shark Detector to facilitate common identification applications. Prediction accuracy of the software pipeline increases as more images are added to the training dataset. We provide public access to the software on our GitHub page.

Keywords: classification, data mining, Instagram, remote monitoring, sharks

Procedia PDF Downloads 122

Authors: Kenji Tanaka, Kazumi Sugita, Yuto Mizushima

Abstract:

This paper introduces a noteworthy form of English communication training for the officers and operators of the Osaka-Bay Marine Traffic Information Service (Osaka MARTIS) of the Japan Coast Guard working in the congested area at the Akashi Strait in Hyogo Prefecture, Japan. The authors of this paper, Marine Technical College’s (MTC) English language instructors, have been holding about forty lectures and exercises in basic and normal Maritime English (ME) for several groups of MARTIS personnel at Osaka MARTIS annually since they started the training in 2005. Trainees are expected to be qualified Maritime Third-Class Radio Operators who are responsible for providing safety information to a daily average of seven to eight hundred vessels that pass through the Akashi Strait, one of Japan’s narrowest channels. As of 2022, the instructors are conducting 55 remote lessons at MARTIS. One lesson is 90 minutes long. All 26 trainees are given oral and written assessments. The trainees need to pass the examination to become qualified operators every year, requiring them to train and maintain their linguistic levels even during the pandemic of Corona Virus Disease-19 (COVID-19). The vessel traffic information provided by Osaka MARTIS in Maritime English language is essential to the work involving the use of very high frequency (VHF) communication between MARTIS and vessels in the area. ME is the common language mainly used on board merchant, fishing, and recreational vessels, normally at sea. ME was edited and recommended by the International Maritime Organization in the 1970s, was revised in 2002, and has undergone continual revision. The vessel’s circumstances are much more serious at the strait than those at the open sea, so these vessels need ME to receive guidance from the center when passing through the narrow strait. The imminent and challenging situations at the strait necessitate that textbooks’ contents include the basics of the phrase book for seafarers as well as specific and additional navigational information, pronunciation exercises, notes on keywords and phrases, explanations about collocations, sample sentences, and explanations about the differences between synonyms especially those focusing on terminologies necessary for passing through the strait. Additionally, short Japanese-English translation quizzes about these topics, as well as prescribed readings about the maritime sector, are include in the textbook. All of these exercises have been trained in the remote education system since the outbreak of COVID-19. According to the guidelines of ME edited in 2009, the lowest level necessary for seafarers is B1 (lower individual users) of The Common European Framework of Reference for Languages: Learning, Teaching, Assessment (CEFR). Therefore, this vocational ME language training at Osaka MARTIS aims for its trainees to communicate at levels higher than B1. A noteworthy proof of improvement from this training is that most of the trainees have become qualified marine radio communication officers.

Keywords: akashi strait, B1 of CEFR, maritime english communication training, osaka martis

Procedia PDF Downloads 124

Authors: Raptis Sotirios

Abstract:

Health and social care (HSc) services planning and scheduling are facing unprecedented challenges due to the pandemic pressure and also suffer from unplanned spending that is negatively impacted by the global financial crisis. Data-driven can help to improve policies, plan and design services provision schedules using algorithms assist healthcare managers’ to face unexpected demands using fewer resources. The paper discusses services packing using statistical significance tests and machine learning (ML) to evaluate demands similarity and coupling. This is achieved by predicting the range of the demand (class) using ML methods such as CART, random forests (RF), and logistic regression (LGR). The significance tests Chi-Squared test and Student test are used on data over a 39 years span for which HSc services data exist for services delivered in Scotland. The demands are probabilistically associated through statistical hypotheses that assume that the target service’s demands are statistically dependent on other demands as a NULL hypothesis. This linkage can be confirmed or not by the data. Complementarily, ML methods are used to linearly predict the above target demands from the statistically found associations and extend the linear dependence of the target’s demand to independent demands forming, thus groups of services. Statistical tests confirm ML couplings making the prediction also statistically meaningful and prove that a target service can be matched reliably to other services, and ML shows these indicated relationships can also be linear ones. Zero paddings were used for missing years records and illustrated better such relationships both for limited years and in the entire span offering long term data visualizations while limited years groups explained how well patients numbers can be related in short periods or can change over time as opposed to behaviors across more years. The prediction performance of the associations is measured using Receiver Operating Characteristic(ROC) AUC and ACC metrics as well as the statistical tests, Chi-Squared and Student. Co-plots and comparison tables for RF, CART, and LGR as well as p-values and Information Exchange(IE), are provided showing the specific behavior of the ML and of the statistical tests and the behavior using different learning ratios. The impact of k-NN and cross-correlation and C-Means first groupings is also studied over limited years and the entire span. It was found that CART was generally behind RF and LGR, but in some interesting cases, LGR reached an AUC=0 falling below CART, while the ACC was as high as 0.912, showing that ML methods can be confused padding or by data irregularities or outliers. On average, 3 linear predictors were sufficient, LGR was found competing RF well, and CART followed with the same performance at higher learning ratios. Services were packed only if when significance level(p-value) of their association coefficient was more than 0.05. Social factors relationships were observed between home care services and treatment of old people, birth weights, alcoholism, drug abuse, and emergency admissions. The work found that different HSc services can be well packed as plans of limited years, across various services sectors, learning configurations, as confirmed using statistical hypotheses.

Keywords: class, cohorts, data frames, grouping, prediction, prob-ability, services

Procedia PDF Downloads 237

Authors: M. Taheri Tehrani, H. Ajorloo

Abstract:

In this paper, an intelligent multi-agent framework is developed for each router in which agents have two vital functionalities, traffic shaping and buffer allocation and are positioned in the ports of the routers. With traffic shaping functionality agents shape the traffic forward by dynamic and real time allocation of the rate of generation of tokens in a Token Bucket algorithm and with buffer allocation functionality agents share their buffer capacity between each other based on their need and the conditions of the network. This dynamic and intelligent framework gives this opportunity to some ports to work better under burst and more busy conditions. These agents work intelligently based on Reinforcement Learning (RL) algorithm and will consider effective parameters in their decision process. As RL have limitation considering much parameter in its decision process due to the volume of calculations, we utilize our novel method which invokes Principle Component Analysis (PCA) on the RL and gives a high dimensional ability to this algorithm to consider as much as needed parameters in its decision process. This implementation when is compared to our previous work where traffic shaping was done without any sharing and dynamic allocation of buffer size for each port, the lower packet drop in the whole network specifically in the source routers can be seen. These methods are implemented in our previous proposed intelligent simulation environment to be able to compare better the performance metrics. The results obtained from this simulation environment show an efficient and dynamic utilization of resources in terms of bandwidth and buffer capacities pre allocated to each port.

Keywords: principal component analysis, reinforcement learning, buffer allocation, multi- agent systems

Procedia PDF Downloads 519

Authors: Despoina Mouratidis, Katia Lida Kermanidis

Abstract:

Machine translation (MT) is a very important task in Natural Language Processing (NLP). MT evaluation is crucial in MT development, as it constitutes the means to assess the success of an MT system, and also helps improve its performance. Several methods have been proposed for the evaluation of (MT) systems. Some of the most popular ones in automatic MT evaluation are score-based, such as the BLEU score, and others are based on lexical similarity or syntactic similarity between the MT outputs and the reference involving higher-level information like part of speech tagging (POS). This paper presents a language-independent machine learning framework for classifying pairwise translations. This framework uses vector representations of two machine-produced translations, one from a statistical machine translation model (SMT) and one from a neural machine translation model (NMT). The vector representations consist of automatically extracted word embeddings and string-like language-independent features. These vector representations used as an input to a multi-layer neural network (NN) that models the similarity between each MT output and the reference, as well as between the two MT outputs. To evaluate the proposed approach, a professional translation and a "ground-truth" annotation are used. The parallel corpora used are English-Greek (EN-GR) and English-Italian (EN-IT), in the educational domain and of informal genres (video lecture subtitles, course forum text, etc.) that are difficult to be reliably translated. They have tested three basic deep learning (DL) architectures to this schema: (i) fully-connected dense, (ii) Convolutional Neural Network (CNN), and (iii) Long Short-Term Memory (LSTM). Experiments show that all tested architectures achieved better results when compared against those of some of the well-known basic approaches, such as Random Forest (RF) and Support Vector Machine (SVM). Better accuracy results are obtained when LSTM layers are used in our schema. In terms of a balance between the results, better accuracy results are obtained when dense layers are used. The reason for this is that the model correctly classifies more sentences of the minority class (SMT). For a more integrated analysis of the accuracy results, a qualitative linguistic analysis is carried out. In this context, problems have been identified about some figures of speech, as the metaphors, or about certain linguistic phenomena, such as per etymology: paronyms. It is quite interesting to find out why all the classifiers led to worse accuracy results in Italian as compared to Greek, taking into account that the linguistic features employed are language independent.

Keywords: machine learning, machine translation evaluation, neural network architecture, pairwise classification

Procedia PDF Downloads 133

Authors: Saikiran Subbagari, Avinash Malladhi

Abstract:

Anti Money Laundering (AML) regulations are designed to prevent money laundering and terrorist financing activities worldwide. Financial institutions around the world are legally obligated to identify, assess and mitigate the risks associated with money laundering and report any suspicious transactions to governing authorities. With increasing volumes of data to analyze, financial institutions seek to automate their AML processes. In the rise of financial crimes, optical character recognition (OCR), in combination with machine learning (ML) algorithms, serves as a crucial tool for automating AML processes by extracting the data from documents and identifying suspicious transactions. In this paper, we examine the utilization of OCR for AML and delve into various OCR techniques employed in AML processes. These techniques encompass template-based, feature-based, neural network-based, natural language processing (NLP), hidden markov models (HMMs), conditional random fields (CRFs), binarizations, pattern matching and stroke width transform (SWT). We evaluate each technique, discussing their strengths and constraints. Also, we emphasize on how OCR can improve the accuracy of customer identity verification by comparing the extracted text with the office of foreign assets control (OFAC) watchlist. We will also discuss how OCR helps to overcome language barriers in AML compliance. We also address the implementation challenges that OCR-based AML systems may face and offer recommendations for financial institutions based on the data from previous research studies, which illustrate the effectiveness of OCR-based AML.

Keywords: anti-money laundering, compliance, financial crimes, fraud detection, machine learning, optical character recognition

Procedia PDF Downloads 146