Search results for: asynchronous input

Authors: S. Jimenez, R.Duarte, J.Sanchez-Choliz, I. Villanua

Abstract:

There is a certain consensus in economic literature about the factors that have influenced in historical differences in growth rates observed between developed and developing countries. However, it is less clear what elements have marked different paths of growth in developed economies in recent decades. R&D has always been seen as one of the major sources of technological progress, and productivity growth, which is directly influenced by technological developments. Following recent literature, we can say that ‘innovation pushes the technological frontier forward’ as well as encourage future innovation through the creation of externalities. In other words, productivity benefits from innovation are not fully appropriated by innovators, but it also spread through the rest of the economies encouraging absorptive capacities, what have become especially important in a context of increasing fragmentation of production This paper aims to contribute to this literature in two ways, first, exploring alternative indexes of R&D flows embodied in inter-country, inter-sectorial flows of good and services (as approximation to technology spillovers) capturing structural and technological characteristic of countries and, second, analyzing the impact of direct and embodied R&D on the evolution of labor productivity at the country/sector level in recent decades. The traditional way of calculation through a multiregional input-output framework assumes that all countries have the same capabilities to absorb technology, but it is not, each one has different structural features and, this implies, different capabilities as part of literature, claim. In order to capture these differences, we propose to use a weight based on specialization structure indexes; one related with the specialization of countries in high-tech sectors and the other one based on a dispersion index. We propose these two measures because, as far as we understood, country capabilities can be captured through different ways; countries specialization in knowledge-intensive sectors, such as Chemicals or Electrical Equipment, or an intermediate technology effort across different sectors. Results suggest the increasing importance of country capabilities while increasing the trade openness. Besides, if we focus in the country rankings, we can observe that with high-tech weighted R&D embodied countries as China, Taiwan and Germany arose the top five despite not having the highest intensities of R&D expenditure, showing the importance of country capabilities. Additionally, through a fixed effects panel data model we show that, in fact, R&D embodied is important to explain labor productivity increases, in fact, even more that direct R&D investments. This is reflecting that globalization is more important than has been said until now. However, it is true that almost all analysis done in relation with that consider the effect of t-1 direct R&D intensity over economic growth. Nevertheless, from our point of view R&D evolve as a delayed flow and it is necessary some time to be able to see its effects on the economy, as some authors have already claimed. Our estimations tend to corroborate this hypothesis obtaining a gap between 4-5 years.

Keywords: economic growth, embodied, input-output, technology

Procedia PDF Downloads 124

Authors: Young-San Shin, Seongsoo Lee

Abstract:

Hall sensor is widely used to measure rotation angle. When the Hall voltage is measured for linear displacement, it is converted to angular displacement using arctangent function, which requires a large lookup table. In this paper, a lookup table reduction technique is presented for angle measurement. When the input of the lookup table is small within a certain threshold, the change of the outputs with respect to the change of the inputs is relatively small. Thus, several inputs can share same output, which significantly reduce the lookup table size. Its error analysis was also performed, and the threshold was determined so as to maintain the error less than 1°. When the Hall voltage has 11-bit resolution, the lookup table size is reduced from 1,024 samples to 279 samples.

Keywords: hall sensor, angle measurement, lookup table, arctangent

Procedia PDF Downloads 337

Authors: Saeedeh Barzegar Khalilsaraei, Manoocheher Kelarestaghi, Farshad Eshghi

Abstract:

The vehicle color, as a prominent and stable feature, helps to identify a vehicle more accurately. As a result, vehicle color recognition is of great importance in intelligent transportation systems. Unlike conventional methods which use only a single Convolutional Neural Network (CNN) for feature extraction or classification, in this paper, four CNNs, with different architectures well-performing in different classes, are trained to extract various features from the input image. To take advantage of the distinct capability of each network, the multiple outputs are combined using a stack generalization algorithm as an ensemble technique. As a result, the final model performs better than each CNN individually in vehicle color identification. The evaluation results in terms of overall average accuracy and accuracy variance show the proposed method’s outperformance compared to the state-of-the-art rivals.

Keywords: Vehicle Color Recognition, Ensemble Algorithm, Stack Generalization, Convolutional Neural Network

Procedia PDF Downloads 85

Authors: Jana Ismail, Layla Fallatah, Maha Alshmaisi

Abstract:

ChatGPT, a cutting-edge language model built on the GPT-3.5 architecture, has garnered attention for its profound natural language processing capabilities, holding promise for transformative applications in customer service and content creation. This study delves into ChatGPT's architecture, aiming to comprehensively understand its strengths and potential limitations. Through systematic experiments across diverse domains, such as general knowledge and creative writing, we evaluated the model's coherence, context retention, and task-specific accuracy. While ChatGPT excels in generating human-like responses and demonstrates adaptability, occasional inaccuracies and sensitivity to input phrasing were observed. The study emphasizes the impact of prompt design on output quality, providing valuable insights for the nuanced deployment of ChatGPT in conversational AI and contributing to the ongoing discourse on the evolving landscape of natural language processing in artificial intelligence.

Keywords: artificial Inelegance, chatGPT, open AI, NLP

Procedia PDF Downloads 77

Authors: Osama Mohamed Elrajubi, Idris El-Feghi, Mohamed Abu Baker Saghayer

Abstract:

This paper describes an identification of specific shapes within binary images using the morphological Hit-or-Miss Transform (HMT). Hit-or-Miss transform is a general binary morphological operation that can be used in searching of particular patterns of foreground and background pixels in an image. It is actually a basic operation of binary morphology since almost all other binary morphological operators are derived from it. The input of this method is a binary image and a structuring element (a template which will be searched in a binary image) while the output is another binary image. In this paper a modification of Hit-or-Miss transform has been proposed. The accuracy of algorithm is adjusted according to the similarity of the template and the sought template. The implementation of this method has been done by C language. The algorithm has been tested on several images and the results have shown that this new method can be used for similar shape detection.

Keywords: hit-or-miss operator transform, HMT, binary morphological operation, shape detection, binary images processing

Procedia PDF Downloads 333

Authors: Jiahe Liu, HongyangYu, Feng Qian, Miao Luo

Abstract:

This study aimed to address the problem of improving the confidence of key points by fusing multi-view information, thereby estimating human posture more accurately. We first obtained multi-view image information and then used the MvP algorithm to fuse this multi-view information together to obtain a set of high-confidence human key points. We used these as the input for the Spatio-Temporal Graph Convolution (ST-GCN). ST-GCN is a deep learning model used for processing spatio-temporal data, which can effectively capture spatio-temporal relationships in video sequences. By using the MvP algorithm to fuse multi-view information and inputting it into the spatio-temporal graph convolution model, this study provides an effective method to improve the accuracy of human posture estimation and provides strong support for further research and application in related fields.

Keywords: multi-view, pose estimation, ST-GCN, joint fusion

Procedia PDF Downloads 70

Authors: Dalvinder Kaur Mangal

Abstract:

For the past decade, noise corrupted output measurements have been a fundamental research problem to be investigated. On the other hand, the estimation of the parameters for linear dynamic systems when also the input is affected by noise is recognized as more difficult problem which only recently has received increasing attention. Representations where errors or measurement noises/disturbances are present on both the inputs and outputs are usually called errors-in-variables (EIV) models. These disturbances may also have additive effects which are also considered in this paper. Parameter estimation of false EIV problem using equation error, output error and iterative prefiltering identification schemes with and without additive uncertainty, when only the output observation is corrupted by noise has been dealt in this paper. The comparative study of these three schemes has also been carried out.

Keywords: errors-in-variable (EIV), false EIV, equation error, output error, iterative prefiltering, Gaussian noise

Procedia PDF Downloads 491

Authors: Ramesh Chandra Majhi

Abstract:

Saturation flow is a very important input variable for the design of signalized intersections. Saturation flow measurement is well established for homogeneous traffic. However, saturation flow measurement and modeling is a challenging task in heterogeneous characterized by multiple vehicle types and non-lane based movement. Present study focuses on proposing a field procedure for Saturation flow measurement and the effect of typical mixed traffic behavior at the signal as far as non-lane based traffic movement is concerned. Data collected during peak and off-peak hour from five intersections with varying approach width is used for validating the saturation flow model. The insights from the study can be used for modeling saturation flow and delay at signalized intersection in heterogeneous traffic conditions.

Keywords: optimization, passenger car unit, saturation flow, signalized intersection

Procedia PDF Downloads 327

Authors: Dong-Bo Hsu

Abstract:

Universal Grammar (UG) claims that the constraints that are derived from this should operate in language users’ L2 grammars. This study investigated this hypothesis on knowledge of Subjacency and resumptive pronoun usage among Chinese learners of English. Chinese fulfills two requirements to examine the existence of UG, i.e., Subjacency does not operate in Chinese and resumptive pronouns in English are very different from those in Chinese and second L2 input undermines the knowledge of Subjacency. The results indicated that Chinese learners of English demonstrated a nearly identical pattern as English native speakers do but the resumptive pronoun in the embedding clauses. This may be explained in terms of the case that Chinese speakers’ usage of pronouns is not influenced by the number of embedding clauses. Chinese learners of English have full access to knowledge endowed by UG but their processing of English sentences may be different from native speakers as a general slow rate for processing in their L2 English.

Keywords: universal grammar, Chinese, English, wh-questions, resumption

Procedia PDF Downloads 469

Authors: Rokhmatul Insani, Hira Laksmiwati Soemitro

Abstract:

Business Intelligence is a methodology that exploits the data to produce information and knowledge systematically, business intelligence can support the decision-making process. Some methods in business intelligence are data warehouse and data mining. A data warehouse can store historical data from transactional data. For data modelling in data warehouse, we apply dimensional modelling by Kimball. While data mining is used to extracting patterns from the data and get insight from the data. Data mining has many techniques, one of which is segmentation. For profiling of telecommunication customer, we use customer segmentation according to customer’s usage of services, customer invoice and customer payment. Customers can be grouped according to their characteristics and can be identified the profitable customers. We apply K-Means Clustering Algorithm for segmentation. The input variable for that algorithm we use RFM (Recency, Frequency and Monetary) model. All process in data mining, we use tools IBM SPSS modeller.

Keywords: business intelligence, customer segmentation, data warehouse, data mining

Procedia PDF Downloads 484

Authors: Nesrine Bouguerra

Abstract:

Entrepreneurship is seen as a necessary ingredient for stimulating economic growth and employment opportunities in all societies. SMEs account for a wide share of economic activity and development. they are the primary engine of job creation, income growth and poverty reduction. Indeed, government support for entrepreneurship is a strategic option to foster economic growth and females’ input in this regard, is of equal significance not only for employability and productivity but also to narrow the gender gap created by social attitudes and beliefs. This study investigates the impact of socio–cultural factors, among other barriers on female entrepreneurial intention in Algeria. Data will be collected using a mixed method approach (Questionnaires and Interviews) from women intending to become entrepreneurs and those already in the field. This study has conceptual, theoretical and empirical contributions to the field of entrepreneurship which will be unveiled throughout.

Keywords: female entrepreneurship, SMEs, women, socio –cultural values, barriers

Procedia PDF Downloads 452

Authors: Kai Chen, Shuguang Cui, Feng Yin

Abstract:

Gaussian process (GP) with spectral mixture (SM) kernel demonstrates flexible non-parametric Bayesian learning ability in modeling unknown function. In this work a novel time-varying and non-stationary convolution spectral mixture (TN-CSM) kernel with a significant enhancing of interpretability by using process convolution is introduced. A way decomposing the SM component into an auto-convolution of base SM component and parameterizing it to be input dependent is outlined. Smoothly, performing a convolution between two base SM component yields a novel structure of non-stationary SM component with much better generalized expression and interpretation. The TN-CSM perfectly allows compatibility with the stationary SM kernel in terms of kernel form and spectral base ignored and confused by previous non-stationary kernels. On synthetic and real-world datatsets, experiments show the time-varying characteristics of hyper-parameters in TN-CSM and compare the learning performance of TN-CSM with popular and representative non-stationary GP.

Keywords: Gaussian process, spectral mixture, non-stationary, convolution

Procedia PDF Downloads 196

Authors: Hamza Nejib, Okba Taouali

Abstract:

This paper presents two of the most knowing kernel adaptive filtering (KAF) approaches, the kernel least mean squares and the kernel recursive least squares, in order to predict a new output of nonlinear signal processing. Both of these methods implement a nonlinear transfer function using kernel methods in a particular space named reproducing kernel Hilbert space (RKHS) where the model is a linear combination of kernel functions applied to transform the observed data from the input space to a high dimensional feature space of vectors, this idea known as the kernel trick. Then KAF is the developing filters in RKHS. We use two nonlinear signal processing problems, Mackey Glass chaotic time series prediction and nonlinear channel equalization to figure the performance of the approaches presented and finally to result which of them is the adapted one.

Keywords: online prediction, KAF, signal processing, RKHS, Kernel methods, KRLS, KLMS

Procedia PDF Downloads 399

Authors: Tarek Saadawi, Rosario Gennaro, Jonathan Akeley

Abstract:

There is rapidly growing demand for professionals to fill positions in Cybersecurity. This is recognized as a national priority both by government agencies and the private sector. Cybersecurity is a very wide technical area which encompasses all measures that can be taken in an electronic system to prevent criminal or unauthorized use of data and resources. This requires defending computers, servers, networks, and their users from any kind of malicious attacks. The need to address this challenge has been recognized globally but is particularly acute in the New York metropolitan area, home to some of the largest financial institutions in the world, which are prime targets of cyberattacks. In New York State alone, there are currently around 57,000 jobs in the Cybersecurity industry, with more than 23,000 unfilled positions. The Cybersecurity Program at City College is a collaboration between the Departments of Computer Science and Electrical Engineering. In Fall 2020, The City College of New York matriculated its first students in theCybersecurity Master of Science program. The program was designed to fill gaps in the previous offerings and evolved out ofan established partnership with Facebook on Cybersecurity Education. City College has designed a program where courses, curricula, syllabi, materials, labs, etc., are developed in cooperation and coordination with industry whenever possible, ensuring that students graduating from the program will have the necessary background to seamlessly segue into industry jobs. The Cybersecurity Program has created multiple pathways for prospective students to obtain the necessary prerequisites to apply in order to build a more diverse student population. The program can also be pursued on a part-time basis which makes it available to working professionals. Since City College’s Cybersecurity M.S. program was established to equip students with the advanced technical skills needed to thrive in a high-demand, rapidly-evolving field, it incorporates a range of pedagogical formats. From its outset, the Cybersecurity program has sought to provide both the theoretical foundations necessary for meaningful work in the field along with labs and applied learning projects aligned with skillsets required by industry. The efforts have involved collaboration with outside organizations and with visiting professors designing new courses on topics such as Adversarial AI, Data Privacy, Secure Cloud Computing, and blockchain. Although the program was initially designed with a single asynchronous course in the curriculum with the rest of the classes designed to be offered in-person, the advent of the COVID-19 pandemic necessitated a move to fullyonline learning. The shift to online learning has provided lessons for future development by providing examples of some inherent advantages to the medium in addition to its drawbacks. This talk will address the structure of the newly-implemented Cybersecurity Master’s Program and discuss the innovations, challenges, and possible future directions.

Keywords: cybersecurity, new york, city college, graduate degree, master of science

Procedia PDF Downloads 147

Authors: Hae-Yeoun Lee

Abstract:

Mosaic refers to a technique that makes image by gathering lots of small materials in various colours. This paper presents an automatic algorithm that makes the photomosaic image using photos. The algorithm is composed of four steps: Partition and feature extraction, block matching, redundancy removal and colour adjustment. The input image is partitioned in the small block to extract feature. Each block is matched to find similar photo in database by comparing similarity with Euclidean difference between blocks. The intensity of the block is adjusted to enhance the similarity of image by replacing the value of light and darkness with that of relevant block. Further, the quality of image is improved by minimizing the redundancy of tiles in the adjacent blocks. Experimental results support that the proposed algorithm is excellent in quantitative analysis and qualitative analysis.

Keywords: photomosaic, Euclidean distance, block matching, intensity adjustment

Procedia PDF Downloads 279

Authors: Wilaiwan Sornpoon, Sébastien Bonnet, Savitri Garivait

Abstract:

Open burning of sugarcane fields is recognized to have a negative impact on soil by degrading its properties, especially soil organic carbon (SOC) content. Better understating the effect of open burning on soil carbon dynamics is crucial for documenting the carbon sequestration capacity of agricultural soils. In this study, experiments to investigate soil carbon stocks under burned and unburned sugarcane plantation systems in Thailand were conducted. The results showed that cultivation fields without open burning during 5 consecutive years enabled to increase the SOC content at a rate of 1.37 Mg ha-1y-1. Also it was found that sugarcane fields burning led to about 15% reduction of the total carbon stock in the 0-30 cm soil layer. The overall increase in SOC under unburned practice is mainly due to the large input of organic material through the use of sugarcane residues.

Keywords: soil organic carbon, soil inorganic carbon, carbon sequestration, open burning, sugarcane

Procedia PDF Downloads 306

Authors: W. S. Barber, S. L. King

Abstract:

This paper reports on a pilot project to develop a collaborative partnership between a community college in rural northern Ontario, Canada, and an urban university in the greater Toronto area in Oshawa, Canada. Partner institutions will collaborate to address learning needs of university applicants whose goals are to attain an undergraduate university BA in Educational Studies and Digital Technology degree, but who may not live in a geographical location that would facilitate this pathways process. The UOIT BA degree is attained through a 2+2 program, where students with a 2 year college diploma or equivalent can attain a four year undergraduate degree. The goals reported on the project are as: 1. Our aim is to expand the BA program to include an additional stream which includes serious educational games, simulations and virtual environments, 2. Develop fully (using both synchronous and asynchronous technologies) online learning modules for use by university applicants who otherwise are not geographically located close to a physical university site, 3. Assess the digital competencies of all students, including members of local, distance and Indigenous communities using a validated tool developed and tested by UOIT across numerous populations. This tool, the General Technical Competency Use and Scale (GTCU) will provide the collaborating institutions with data that will allow for analyzing how well students are prepared to succeed in fully online learning communities. Philosophically, the UOIT BA program is based on a fully online learning communities model (FOLC) that can be accessed from anywhere in the world through digital learning environments via audio video conferencing tools such as Adobe Connect. It also follows models of adult learning and mobile learning, and makes a university degree accessible to the increasing demographic of adult learners who may use mobile devices to learn anywhere anytime. The program is based on key principles of Problem Based Learning, allowing students to build their own understandings through the co-design of the learning environment in collaboration with the instructors and their peers. In this way, this degree allows students to personalize and individualize the learning based on their own culture, background and professional/personal experiences. Using modified flipped classroom strategies, students are able to interrogate video modules on their own time in preparation for one hour discussions occurring in video conferencing sessions. As a consequence of the program flexibility, students may continue to work full or part time. All of the partner institutions will co-develop four new modules, administer the GTCU and share data, while creating a new stream of the UOIT BA degree. This will increase accessibility for students to bridge from community colleges to university through a fully digital environment. We aim to work collaboratively with Indigenous elders, community members and distance education instructors to increase opportunities for more students to attain a university education.

Keywords: aboriginal, college, competencies, digital, universities

Procedia PDF Downloads 215

Authors: Pocholo Rodriguez, Anne Bernadine Ocampo, Ian Benedict Chan, Janric Micah Gray

Abstract:

The state of an electric power system may be classified as either stable or unstable. The borderline of stability is at any condition for which a slight change in an unfavourable direction of any pertinent quantity will cause instability. Voltage instability in power distribution systems could lead to voltage collapse and thus power blackouts. The researchers will present an intelligent system using back propagation algorithm that can detect voltage instability and output voltage of a power distribution and classify it as stable or unstable. The researchers’ work is the use of parameters involved in voltage instability as input parameters to the neural network for training and testing purposes that can provide faster detection and monitoring of the power distribution system.

Keywords: back-propagation algorithm, load instability, neural network, power distribution system

Procedia PDF Downloads 435

Authors: Sachin Nagargoje

Abstract:

Complete labeled data is often difficult to obtain in a practical scenario. Even if one manages to obtain the data, the quality of the data is always in question. In shopping vertical, offers are the input data, which is given by advertiser with or without a good quality of information. In this paper, an author investigated the possibility of using a very simple Semi-supervised learning approach to increase the recall of unhealthy offers (has badly written Offer Title or partial product details) in shopping vertical domain. The author found that the semisupervised learning method had improved the recall in the Smart Phone category by 30% on A=B testing on 10% traffic and increased the YoY (Year over Year) number of impressions per month by 33% at production. This also made a significant increase in Revenue, but that cannot be publicly disclosed.

Keywords: semi-supervised learning, clustering, recall, coverage

Procedia PDF Downloads 122

Authors: Yixin Zhu, Lishengsa Yue, Jian Sun, Lanyue Tang

Abstract:

Modeling driving distraction in microscopic traffic simulation is crucial for enhancing simulation accuracy. Current driving distraction models are mainly derived from physical motion constraints under distracted states, in which distraction-related error terms are added to existing microscopic driver models. However, the model accuracy is not very satisfying, due to a lack of modeling the cognitive mechanism underlying the distraction. This study models driving distraction based on the Queueing Network Human Processor model (QN-MHP). This study utilizes the queuing structure of the model to perform task invocation and switching for distracted operation and control of the vehicle under driver distraction. Based on the assumption of the QN-MHP model about the cognitive sub-network, server F is a structural bottleneck. The latter information must wait for the previous information to leave server F before it can be processed in server F. Therefore, the waiting time for task switching needs to be calculated. Since the QN-MHP model has different information processing paths for auditory information and visual information, this study divides driving distraction into two types: auditory distraction and visual distraction. For visual distraction, both the visual distraction task and the driving task need to go through the visual perception sub-network, and the stimuli of the two are asynchronous, which is called stimulus on asynchrony (SOA), so when calculating the waiting time for switching tasks, it is necessary to consider it. In the case of auditory distraction, the auditory distraction task and the driving task do not need to compete for the server resources of the perceptual sub-network, and their stimuli can be synchronized without considering the time difference in receiving the stimuli. According to the Theory of Planned Behavior for drivers (TPB), this study uses risk entropy as the decision criterion for driver task switching. A logistic regression model is used with risk entropy as the independent variable to determine whether the driver performs a distraction task, to explain the relationship between perceived risk and distraction. Furthermore, to model a driver’s perception characteristics, a neurophysiological model of visual distraction tasks is incorporated into the QN-MHP, and executes the classical Intelligent Driver Model. The proposed driving distraction model integrates the psychological cognitive process of a driver with the physical motion characteristics, resulting in both high accuracy and interpretability. This paper uses 773 segments of distracted car-following in Shanghai Naturalistic Driving Study data (SH-NDS) to classify the patterns of distracted behavior on different road facilities and obtains three types of distraction patterns: numbness, delay, and aggressiveness. The model was calibrated and verified by simulation. The results indicate that the model can effectively simulate the distracted car-following behavior of different patterns on various roadway facilities, and its performance is better than the traditional IDM model with distraction-related error terms. The proposed model overcomes the limitations of physical-constraints-based models in replicating dangerous driving behaviors, and internal characteristics of an individual. Moreover, the model is demonstrated to effectively generate more dangerous distracted driving scenarios, which can be used to construct high-value automated driving test scenarios.

Keywords: computational cognitive model, driving distraction, microscopic traffic simulation, psychological-physical constraints

Procedia PDF Downloads 91

Authors: Koray U. Erbas

Abstract:

Neural networks are appealing for many applications since they are able to learn complex non-linear relationships between input and output data. As the number of neurons and layers in a neural network increase, it is possible to represent more complex relationships with automatically extracted features. Nowadays Deep Neural Networks (DNNs) are widely used in Computer Vision problems such as; classification, object detection, segmentation image editing etc. In this work, Facial Emotion Recognition task is performed by proposed Convolutional Neural Network (CNN)-based DNN architecture using FER2013 Dataset. Moreover, the effects of different hyperparameters (activation function, kernel size, initializer, batch size and network size) are investigated and ablation study results for Pooling Layer, Dropout and Batch Normalization are presented.

Keywords: convolutional neural network, deep learning, deep learning based FER, facial emotion recognition

Procedia PDF Downloads 274

Authors: R. K. Chaurasiya, N. D. Londhe, S. Ghosh

Abstract:

The study of the electrical signals produced by neural activities of human brain is called Electroencephalography. In this paper, we propose an automatic and efficient EEG signal classification approach. The proposed approach is used to classify the EEG signal into two classes: epileptic seizure or not. In the proposed approach, we start with extracting the features by applying Discrete Wavelet Transform (DWT) in order to decompose the EEG signals into sub-bands. These features, extracted from details and approximation coefficients of DWT sub-bands, are used as input to Principal Component Analysis (PCA). The classification is based on reducing the feature dimension using PCA and deriving the support-vectors using Support Vector Machine (SVM). The experimental are performed on real and standard dataset. A very high level of classification accuracy is obtained in the result of classification.

Keywords: discrete wavelet transform, electroencephalogram, pattern recognition, principal component analysis, support vector machine

Procedia PDF Downloads 639

Authors: Sikiru Ademola Adewale, Joe Thomas, Bolanle Hafiz Matti, Tosin Ige

Abstract:

This project demonstrates the implementation and use of an encoder-decoder model to perform a many-to-many mapping of video data to text captions. The many-to-many mapping occurs via an input temporal sequence of video frames to an output sequence of words to form a caption sentence. Data preprocessing, model construction, and model training are discussed. Caption correctness is evaluated using 2-gram BLEU scores across the different splits of the dataset. Specific examples of output captions were shown to demonstrate model generality over the video temporal dimension. Predicted captions were shown to generalize over video action, even in instances where the video scene changed dramatically. Model architecture changes are discussed to improve sentence grammar and correctness.

Keywords: decoder, encoder, many-to-many mapping, video captioning, 2-gram BLEU

Procedia PDF Downloads 108

Authors: Denis Özdemir, Max Senges

Abstract:

There are many examples of AI-supported apps for better health and wellbeing. Generally, these applications help people to achieve their goals based on scientific research and input data. Still, they do not always explain how those three are related, e.g. by making implicit assumptions about goals that hold for many but not for all. We present a seven-step method for designing health and wellbeing AIs considering goal setting, measurable results, real-time indicators, analytics, visual representations, communication, and feedback. It can help engineers as guidance in developing apps, recommendation algorithms, and interfaces that support humans in their decision-making without patronization. To illustrate the method, we create a recommender AI for tiny wellbeing habits and run a small case study, including a survey. From the results, we infer how people perceive the relationship between them and the AI and to what extent it helps them to achieve their goals. We review our seven-step engineering method and suggest modifications for the next iteration.

Keywords: recommender systems, natural language processing, health apps, engineering methods

Procedia PDF Downloads 165

Authors: Yildiz Stella Dak, Jale Tezcan

Abstract:

Ground motion models that relate a strong motion parameter of interest to a set of predictive seismological variables describing the earthquake source, the propagation path of the seismic wave, and the local site conditions constitute a critical component of seismic hazard analyses. When a sufficient number of strong motion records are available, ground motion relations are developed using statistical analysis of the recorded ground motion data. In regions lacking a sufficient number of recordings, a synthetic database is developed using stochastic, theoretical or hybrid approaches. Regardless of the manner the database was developed, ground motion relations are developed using regression analysis. Development of a ground motion relation is a challenging process which inevitably requires the modeler to make subjective decisions regarding the inclusion criteria of the recordings, the functional form of the model and the set of seismological variables to be included in the model. Because these decisions are critically important to the validity and the applicability of the model, there is a continuous interest on procedures that will facilitate the development of ground motion models. This paper proposes the use of the Least Absolute Shrinkage and Selection Operator (LASSO) in selecting the set predictive seismological variables to be used in developing a ground motion relation. The LASSO can be described as a penalized regression technique with a built-in capability of variable selection. Similar to the ridge regression, the LASSO is based on the idea of shrinking the regression coefficients to reduce the variance of the model. Unlike ridge regression, where the coefficients are shrunk but never set equal to zero, the LASSO sets some of the coefficients exactly to zero, effectively performing variable selection. Given a set of candidate input variables and the output variable of interest, LASSO allows ranking the input variables in terms of their relative importance, thereby facilitating the selection of the set of variables to be included in the model. Because the risk of overfitting increases as the ratio of the number of predictors to the number of recordings increases, selection of a compact set of variables is important in cases where a small number of recordings are available. In addition, identification of a small set of variables can improve the interpretability of the resulting model, especially when there is a large number of candidate predictors. A practical application of the proposed approach is presented, using more than 600 recordings from the National Geospatial-Intelligence Agency (NGA) database, where the effect of a set of seismological predictors on the 5% damped maximum direction spectral acceleration is investigated. The set of candidate predictors considered are Magnitude, Rrup, Vs30. Using LASSO, the relative importance of the candidate predictors has been ranked. Regression models with increasing levels of complexity were constructed using one, two, three, and four best predictors, and the models’ ability to explain the observed variance in the target variable have been compared. The bias-variance trade-off in the context of model selection is discussed.

Keywords: ground motion modeling, least absolute shrinkage and selection operator, penalized regression, variable selection

Procedia PDF Downloads 330

Authors: Juan P. Silva, T. R. Canchala, H. J. Lubberding, E. J. Peña, H. J. Gijzen

Abstract:

This study measured the fluxes of greenhouse gases (GHGs) i.e. CO₂, CH₄ and N₂O from a tropical eutrophic freshwater wetland (“Sonso Lagoon”) which receives input loading nutrient from several sources i.e. agricultural run-off, domestic sewage, and a polluted river. The flux measurements were carried out at four different points using the static chamber technique. CO₂ fluxes ranged from -8270 to 12210 mg.m^-2.d^-1 (median = 360; SD = 4.11; n = 50), CH₄ ranged between 0.2 and 5270 mg.m^-2.d^-1 (median = 60; SD = 1.27; n = 45), and N₂O ranged from -31.12 to 15.4 mg N₂O m^-2.d^-1 (median = 0.05; SD = 9.36; n = 42). Although some negative fluxes were observed in the zone dominated by floating plants i.e. Eichornia crassipes, Salvinia sp., and Pistia stratiotes L., the mean values indicated that the Sonso Lagoon was a net source of CO₂, CH₄ and N₂O. In addition, an effect of the eutrophication on GHG emissions could be observed in the positive correlation found between CO₂, CH₄ and N₂O generation and COD, PO₄^-3, NH₃-N, TN and NO₃^-N. The eutrophication impact on GHG production highlights the necessity to limit the anthropic activities on freshwater wetlands.

Keywords: eutrophication, greenhouse gas emissions, freshwater wetlands, climate change

Procedia PDF Downloads 361

Authors: S. J. Ha, S. J. Baik, T. O. Kim, S. W. Han

Abstract:

For computing the seismic responses of structures, current seismic design provisions permit response history analyses (RHA) that can be used without limitations in height, seismic design category, and building irregularity. In order to obtain accurate seismic responses using RHA, it is important to use adequate input ground motions. Current seismic design provisions provide criteria for selecting ground motions. In this study, the accurate and computationally efficient algorithm is proposed for accurately selecting ground motions that satisfy the requirements specified in current seismic design provisions. The accuracy of the proposed algorithm is verified using single-degree-of-freedom systems with various natural periods and yield strengths. This study shows that the mean seismic responses obtained from RHA with seven and ten ground motions selected using the proposed algorithm produce errors within 20% and 13%, respectively.

Keywords: algorithm, ground motion, response history analysis, selection

Procedia PDF Downloads 286

Authors: Katerina Samartzi

Abstract:

This study focuses on the teacher’s discourses in Greece, about the relation between the children and the media, concerning sex education and widely the childhood. The teachers’ input reflect the anxieties and the dominant discourses that exist around these issues. The study begins with the critical discussion of the available literature concerning the potential impact of media and the ‘moral panics’, their role in sex education and the children’s use of sexual material. Moreover, the study analyses the social construction of childhood and sexuality. Given the lack of explicit and official protocol for the sex education in Greece and due the fact that the young people are familiar with all the material provided by the New Media and their part as an informal education, this project aims to point out the factors that reinforce these gaps. This study focuses on the way the adults and specifically teachers contextualize the children’s relation with media, their sexuality, the sex education, the use of sexual material and the childhood.

Keywords: childhood, children's sexuality, media, moral panics, pornography, sex education

Procedia PDF Downloads 175

Authors: Jiewei Li, Hongyan Cui, Chunqi Chang, Yong Hu

Abstract:

It was expected to benefit patient with hemiparesis after stroke by extensive arm rehabilitation, to partially regain forearm and hand function. This paper propose a robotic rehabilitation arm in assisting the hemiparetic patient to learn new ways of using and moving their weak arms. In this study, the robotic arm was driven by a somatosensory stimulated brain computer interface (BCI), which is a new modality BCI. The use of somatosensory stimulation is not only an input for BCI, but also a electrical stimulation for treatment of hemiparesis to strengthen the arm and improve its range of motion. A trial of this robotic rehabilitation arm was performed in a stroke patient with pure motor hemiparesis. The initial trial showed a promising result from the patient with great motivation and function improvement. It suggests that robotic rehabilitation arm driven by somatosensory BCI can enhance the rehabilitation performance and progress for hemiparetic patients after stroke.

Keywords: robotic rehabilitation arm, brain computer interface (BCI), hemiparesis, stroke, somatosensory stimulation

Procedia PDF Downloads 390

Authors: Armine Vahanyan

Abstract:

Humanity has been still coping with the new COVID-19 pandemic. Healthcare providers, economists, psychologists, and other specialists speak about the impact of the virus on different spheres of our life. In the list of similar discussions, the impact of pandemics on global education is of utmost importance. Ideally, providing quality education services should be crucial, and the ways education programs are being adapted will determine the success or failure of the service providers. The paper aims to summarize the research touching upon the current situation of higher education in Armenia. The research includes data from official reports, surveys among education leads, faculty, and students, as well as personal observations and consideration. Through descriptive analysis, the findings of the research are being presented from various aspects. Interim results of the research unveiled two major issues in the sector of higher education in Armenia. On the one hand, the entire compulsory digitization of instruction, assessment, and grading has evoked serious gaps related to the lack of technical competencies. There is an urgent need for professional development programs that will address most of the concerns due to the shift to the online instruction mode. On the other hand, online teaching and learning require revision and adaptation of the existing curricula. Given that the content of certain programs may not be compromised, the teaching methods, the assignments, and evaluation require profound transformation, which will still be in line with course learning outcomes and student learning outcomes. The given paper focuses on the ways the mentioned issues are being addressed in Armenia. The extent of commitment for changes and adaptability to the new situation varies from the government-funded and private universities. In particular, the paper compares and contrasts activities and measures taken at the Armenian State Pedagogical University and the American University of Armenia. Thus, the Pedagogical University focused on the use of Google Classroom as the only means for teaching and learning as well as adopted the compulsory synchronous instruction mode. The American University, on the contrary, kept practicing the academic freedom, enabling both synchronous and asynchronous instruction modes, ensuring alignment of the course learning outcomes and student learning outcomes. The State University utilized the assignments and assessment, which would work for the on-campus instruction mode, while the American university employed a variety of assignments applicable for online teaching mode. The latter has suggested the utilization of multiple apps, internet sources, and online library access for a better online instant. Discussions with faculty through online forums and/or professional development workshops also facilitate restructuring and adaptation of the courses. Finally, the paper will synthesize the results of the undertaken research and will outline the e-learning perspectives and opportunities boosted by the known devastating healthcare issue.

Keywords: assessment, compulsory digitization of education services, online teaching, instruction mode, program restructuring

Procedia PDF Downloads 127