Search results for: word learning

Authors: Belkhadir Radouane

Abstract:

In this study, author tried to explore the beliefs of some Moroccan teachers working in the delegations of Safi and Youcefia about the usefulness of first and second languages in learning the third language. More specifically, author attempted to see the extent to which these teachers believe that a first and second language can serve students in learning a third one. The first language in this context is Arabic, the second is French, and the third is English. The teachers’ beliefs were gathered through a questionnaire that was addressed via Google Forms. Then, the results were analyzed using the same application. It was found that teachers are positive about the usefulness of the first and second language in learning the third one, but most of them rarely use in a conscious way activities that serve this purpose.

Keywords: Bilinguilism, teachers beliefs, English as ESL, Morocco

Procedia PDF Downloads 56

Authors: Gayathri Sadanala, Shibam Pokhrel, Owen Murphy

Abstract:

In the ever-evolving landscape of education, Virtual Reality (VR) environments offer a promising avenue for enhancing student engagement and learning experiences. However, understanding and predicting student behavior within these immersive settings remain challenging tasks. This paper presents a comprehensive study on the predictive modeling of student behavior in VR using machine learning techniques. We introduce a rich data set capturing student interactions, movements, and progress within a VR orientation program. The dataset is divided into training and testing sets, allowing us to develop and evaluate predictive models for various aspects of student behavior, including engagement levels, task completion, and performance. Our machine learning approach leverages a combination of feature engineering and model selection to reveal hidden patterns in the data. We employ regression and classification models to predict student outcomes, and the results showcase promising accuracy in forecasting behavior within VR environments. Furthermore, we demonstrate the practical implications of our predictive models for personalized VR-based learning experiences and early intervention strategies. By uncovering the intricate relationship between student behavior and VR interactions, we provide valuable insights for educators, designers, and developers seeking to optimize virtual learning environments.

Keywords: interaction, machine learning, predictive modeling, virtual reality

Procedia PDF Downloads 144

Authors: Megha Gupta, Nupur Prakash

Abstract:

Identification of plant diseases has been performed using machine learning and deep learning models on the datasets containing images of healthy and diseased plant leaves. The current study carries out an evaluation of some of the deep learning models based on convolutional neural network (CNN) architectures for identification of plant diseases. For this purpose, the publicly available New Plant Diseases Dataset, an augmented version of PlantVillage dataset, available on Kaggle platform, containing 87,900 images has been used. The dataset contained images of 26 diseases of 14 different plants and images of 12 healthy plants. The CNN models selected for the study presented in this paper are AlexNet, ZFNet, VGGNet (four models), GoogLeNet, and ResNet (three models). The selected models are trained using PyTorch, an open-source machine learning library, on Google Colaboratory. A comparative study has been carried out to analyze the high degree of accuracy achieved using these models. The highest test accuracy and F1-score of 99.59% and 0.996, respectively, were achieved by using GoogLeNet with Mini-batch momentum based gradient descent learning algorithm.

Keywords: comparative analysis, convolutional neural networks, deep learning, plant disease identification

Procedia PDF Downloads 199

Authors: Pavan K. Rallabandi, Kailash C. Patidar

Abstract:

In this paper, we present an optimization technique or a learning algorithm using the hybrid architecture by combining the most popular sequence recognition models such as Recurrent Neural Networks (RNNs) and Hidden Markov models (HMMs). In order to improve the sequence or pattern recognition/ classification performance by applying a hybrid/neural symbolic approach, a gradient descent learning algorithm is developed using the Real Time Recurrent Learning of Recurrent Neural Network for processing the knowledge represented in trained Hidden Markov Models. The developed hybrid algorithm is implemented on automata theory as a sample test beds and the performance of the designed algorithm is demonstrated and evaluated on learning the deterministic finite state automata.

Keywords: hybrid systems, hidden markov models, recurrent neural networks, deterministic finite state automata

Procedia PDF Downloads 388

Authors: Eirini Busack

Abstract:

Since the outbreak of the Covid-19 pandemic and the sudden transfer to online teaching, teachers have struggled to reconstruct their teaching and learning materials to adapt them to the new reality of online teaching and learning. Consequently, the pupils’ learning was disrupted during this orientation phase. Due to the above situation, teachers from all fields concluded that it is vital that their pupils should be able to continue their learning even without the teacher being physically present. Various websites and applications have been in use since then in hope that pupils will still enjoy a qualitative education; unfortunately, this was often not the case. To address this issue, it was therefore decided to focus the research on the development of digital learning paths. The fundamentals of these learning paths include the implementation of scenario-based learning (digital storytelling), the integration of media-didactic theory to make it pedagogically appropriate for learners, alongside instructional design knowledge and the drive to promote autonomous learners. This particular research is being conducted within the frame of the research project “Sustainable integration of subject didactic digital teaching-learning concepts” (InDiKo, 2020-2023), which is currently conducted at the University of Education Karlsruhe and investigates how pre-service teachers can acquire the necessary interdisciplinary and subject-specific media-didactic competencies to provide their future learners with digitally enhanced learning opportunities, and how these competencies can be developed continuously and sustainably. As English is one of the subjects involved in this project, the English Department prepared a seminar for the pre-service secondary teachers: “Media-didactic competence development: Developing learning paths & Digital Storytelling for English grammar teaching.” During this seminar, the pre-service teachers plan and design a Moodle-based differentiated lesson sequence on an English grammar topic that is to be tested by secondary school pupils. The focus of the present research is to assess the secondary school pupils’ expectations from an English grammar-focused digital learning path created by pre-service English teachers. The nine digital learning paths that are to be distributed to 25 pupils were produced over the winter and the current summer semester as the artifact of the seminar. Finally, the data to be quantitatively analysed and interpreted derive from the online questionnaires that the secondary school pupils fill in so as to reveal their expectations on what they perceive as a stimulating and thus effective grammar-focused digital learning path.

Keywords: digital storytelling, learning paths, media-didactics, autonomous learning

Procedia PDF Downloads 81

Authors: Vaida Jurgile

Abstract:

Intercultural learning is one of the approaches taken to understand the cultural diversity of the modern world and to accept changes in cultural identity and otherness and the expression of tolerance. During intercultural learning, students develop their abilities to interact and communicate with their group members. These abilities help to understand social and cultural differences, to form one’s identity, and to give meaning to intercultural learning. Intercultural education recognizes that a true understanding of differences and similarities of another culture is necessary in order to lay the foundations for working together with others, which contributes to the promotion of intercultural dialogue, appreciation of diversity, and cultural exchange. Therefore, it is important to examine the concept of intercultural learning, revealed through students’ learning experiences and understanding of how this learning takes place and what significance this phenomenon has in higher education. At a scientific level, intercultural learning should be explored in order to uncover the influence of cultural identity, i.e., intercultural learning should be seen in a local context. This experience would provide an opportunity to learn from various everyday intercultural learning situations. Intercultural learning can be not only a form of learning but also a tool for building understanding between people of different cultures. The research object of the study is the process of intercultural learning. The aim of the dissertation is to develop a grounded theory of the process of learning in an intercultural study environment, revealing students’ learning experiences. The research strategy chosen in this study is a constructivist grounded theory (GT). GT is an inductive method that seeks to form a theory by applying the systematic collection, synthesis, analysis, and conceptualization of data. The targeted data collection was based on the analysis of data provided by previous research participants, which revealed the need for further research participants. During the research, only students with at least half a year of study experience, i.e., who have completed at least one semester of intercultural studies, were purposefully selected for the research. To select students, snowballing sampling was used. 18 interviews were conducted with students representing 3 different fields of sciences (social sciences, humanities, and technology sciences). In the process of intercultural learning, language expresses and embodies cultural reality and a person’s cultural identity. It is through language that individual experiences are expressed, and the world in which Others exist is perceived. The increased emphasis is placed on the fact that language conveys certain “signs’ of communication and perception with cultural value, enabling the students to identify the Self and the Other. Language becomes an important tool in the process of intercultural communication because it is only through language that learners can communicate, exchange information, and understand each other. Thus, in the process of intercultural learning, language either promotes interpersonal relationships with foreign students or leads to mutual rejection.

Keywords: intercultural learning, grounded theory, students, other

Procedia PDF Downloads 67

Authors: Geraldine Macdonald, Oliver Perra, Nina O’Neill, Laura Neeson, Kathryn Higgins

Abstract:

Background: Despite improvements in recent years, almost one in six children in Northern Ireland (NI) leaves primary school without achieving the expected level in English and Maths. By early adolescence, this ratio is one in five. In 2010-11, around 9000 pupils in NI had failed to achieve the required standard in literacy and numeracy by the time they left full-time education. This paper reports the findings of an experimental evaluation of a programmed designed to improve educational outcomes of a cohort of children starting primary school in areas of high social disadvantage in Northern Ireland. The intervention: ‘Ready to Learn’ comprised two key components: a literacy-rich After School programme (one hour after school, three days per week), and a range of activities and support to promote the engagement of parents with their children’s learning, in school and at home. The intervention was delivered between September 2010 and August 2013. Study aims and objectives: The primary aim was to assess whether, and to what extent, ‘Ready to Learn’ improved the literacy of socially disadvantaged children entering primary schools compared with children in schools without access to the programme. Secondary aims included assessing the programme’s impact on children’s social, emotional and behavioural regulation, and parents’ engagement with their children’s learning. In total, 505 children (almost all) participated in the baseline assessment for the study, with good retention over seven sweeps of data collection. Study design: The intervention was evaluated by means of a cluster randomized trial, with schools as the unit of randomization and analysis. It included a qualitative component designed to examine process and implementation, and to explore the concept of parental engagement. Sixteen schools participated, with nine randomized to the experimental group. As well as outcome data relating to children, 134 semi-structured interviews were conducted with parents over the three years of the study, together with 88 interviews with school staff. Results: Given the children’s ages, not all measures used were direct measures of reading. Findings point to a positive impact of “Ready to Learn” on children’s reading achievement (comprehension and fluency), as assessed by the York Assessment of Reading Comprehension (YARC) and decoding, assessed using the Word Recognition and Phonic Skills (WRaPS3). Effects were not large, but evidence suggests that it is unusual for an after school programme to clearly to demonstrate effects on reading skills. No differences were found on three other measures of literacy-related skills: British Picture Vocabulary Scale (BPVS-II), Naming Speed and Non-word Reading Tests from the Phonological Assessment Battery (PhAB) or Concepts about Print (CAP) – the last due to an age-related ceiling effect). No differences were found between the two groups on measures of social, emotional and behavioural regulation, and due to low levels of participation, it was not possible directly to assess the contribution of the parent component to children’s outcomes. The qualitative data highlighted conflicting concepts of engagement between parents and school staff. Ready to Learn is a promising intervention that merits further support and evaluation.

Keywords: after-school, education, literacy, parental engagement

Procedia PDF Downloads 379

Authors: David S. Rosenstein

Abstract:

The author believes that modern spoken languages should be equally difficult (or easy) to learn, since all normal children learning their native languages do so at approximately the same rate and with the same competence, progressing from easy to more complex grammar and syntax in the same way. Why then, do some languages seem more difficult than others? Perhaps people are referring to the written language, where it may be true that mastering Chinese requires more time than French, which in turn requires more time than Spanish. But this may be marginal, since Chinese and French children quickly catch up to their Spanish peers in reading comprehension. Rather, the real differences in difficulty derive from two sources: hardened L1 language habits trying to cope with contrasting L2 habits; and unfamiliarity with unique L2 characteristics causing faulty expectations. It would seem that effective L2 teaching and learning must take these two sources of difficulty into consideration. The author feels that the latter (faulty expectations) causes the greatest difficulty, making effective teaching and learning somewhat different for each given foreign language. Examples from Chinese and other languages are presented.

Keywords: learning different languages, language learning difficulties, faulty language expectations

Procedia PDF Downloads 533

Authors: M. Leonora D. Guerrero

Abstract:

The study attempted to uncover the most and least frequent phonological variation evident in the speech patterns of grade 1 teachers in select public elementary schools in the Philippines. It also determined the lectal description of the participants based on Tayao’s consonant charts for American and Philippine English. Descriptive method was utilized. A total of 24 grade 1 teachers participated in the study. The instrument used was word list. Each column in the word list is represented by words with the target consonant phonemes: labiodental fricatives f/ and /v/ and lingua-alveolar fricative /z/. These phonemes were in the initial, medial, and final positions, respectively. Findings of the study revealed that the most frequent variation happened when the participants read words with /z/ in the final position while the least frequent variation happened when the participants read words with /z/ in the initial position. The study likewise proved that the grade 1 teachers exhibited the segmental features of both the mesolect and basilect. Based on these results, it is suggested that teachers of English in the Philippines must aspire to manifest the features of the mesolect, if not, the acrolect since it is expected of the academicians not to be displaying the phonological features of the acrolects since this variety is only used by the 'uneducated.' This is especially so with grade 1 teachers who are often mimicked by their students who classify their speech as the 'standard.'

Keywords: consonant phonemes, lectal description, Philippine English, phonological variation

Procedia PDF Downloads 214

Authors: Leidy M. Aldana, Jorge E. Camargo

Abstract:

Adversarial Machine Learning has gained importance in recent years as Cybersecurity has gained too, especially malware, it has affected different entities and people in recent years. This paper shows a literature review about defense methods created to prevent adversarial machine learning attacks, firstable it shows an introduction about the context and the description of some terms, in the results section some of the attacks are described, focusing on detecting adversarial examples before coming to the machine learning algorithm and showing other categories that exist in defense. A method with five steps is proposed in the method section in order to define a way to make the literature review; in addition, this paper summarizes the contributions in this research field in the last seven years to identify research directions in this area. About the findings, the category with least quantity of challenges in defense is the Detection of adversarial examples being this one a viable research route with the adaptive approach in attack and defense.

Keywords: Malware, adversarial, machine learning, defense, attack

Procedia PDF Downloads 63

Authors: Gihan Thilakarathna, Fernando Ishara, Rathnayake Yasith, Bandara A. M. R. Y.

Abstract:

E-learning is a crucial component of intelligent education. Even in the midst of a pandemic, E-learning is becoming increasingly important in the educational system. Several e-learning programs are accessible for students. Here, we decided to create an e-learning framework for children. We've found a few issues that teachers are having with their online classes. When there are numerous students in an online classroom, how does a teacher recognize a student's focus on academics and below-the-surface behaviors? Some kids are not paying attention in class, and others are napping. The teacher is unable to keep track of each and every student. Key challenge in e-learning is online exams. Because students can cheat easily during online exams. Hence there is need of exam proctoring is occurred. In here we propose an automated online exam cheating detection method using a web camera. The purpose of this project is to present an E-learning platform for math education and include games for kids as an alternative teaching method for math students. The game will be accessible via a web browser. The imagery in the game is drawn in a cartoonish style. This will help students learn math through games. Everything in this day and age is moving towards automation. However, automatic answer evaluation is only available for MCQ-based questions. As a result, the checker has a difficult time evaluating the theory solution. The current system requires more manpower and takes a long time to evaluate responses. It's also possible to mark two identical responses differently and receive two different grades. As a result, this application employs machine learning techniques to provide an automatic evaluation of subjective responses based on the keyword provided to the computer as student input, resulting in a fair distribution of marks. In addition, it will save time and manpower. We used deep learning, machine learning, image processing and natural language technologies to develop these research components.

Keywords: math, education games, e-learning platform, artificial intelligence

Procedia PDF Downloads 156

Authors: Matthias Grünke

Abstract:

In this single-case study, we evaluated the effects of a self-graphing intervention to help students improve their reading fluency. Our participant was a 10-year-old girl with a suspected learning disability in reading. We applied an ABAB reversal design to test the efficacy of our approach. The dependent measure was the number of correctly read words from a children’s book within five minutes. Our participant recorded her daily performance using a simple line diagram. Results indicate that her reading rate improved simultaneously with the intervention and dropped as soon as the treatment was suspended. The findings give reasons for optimism that our simple strategy can be a very effective tool in supporting students with learning disabilities to boost their reading fluency.

Keywords: single-case study, learning disabilities, elementary education, reading problems, reading fluency

Procedia PDF Downloads 112

Authors: Enrique Barra, Aldo Gordillo, Juan Quemada

Abstract:

This paper presents the Virtual Science Hub platform. It is an open source platform that combines a social network, an e-learning authoring tool, a video conference service and a learning object repository for science teaching enrichment. These four main functionalities fit very well together. The platform was released in April 2012 and since then it has not stopped growing. Finally we present the results of the surveys conducted and the statistics gathered to validate this approach.

Keywords: e-learning, platform, authoring tool, science teaching, educational sciences

Procedia PDF Downloads 397

Authors: Suriyani Ariffin, Nor Aziah Alias, Khairil Iskandar Othman, Haslinda Yusoff

Abstract:

An Integrated Cumulative Grade Point Average (iCGPA) is a mechanism and strategy to ensure the curriculum of an academic programme is constructively aligned to the expected learning outcomes and student performance based on the attainment of those learning outcomes that is reported objectively in a spider web. Much effort and time has been spent to develop a viable mechanism and trains academics to utilize the platform for reporting. The question is: How well do learners conceive the idea of their achievement via iCGPA and whether quality learner attributes have been nurtured through the iCGPA mechanism? This paper presents the architecture of an integrated CGPA mechanism purported to address a holistic evaluation from the evaluation of courses learning outcomes to aligned programme learning outcomes attainment. The paper then discusses the students’ understanding of the mechanism and evaluation of their achievement from the generated spider web. A set of questionnaires were distributed to a group of students with iCGPA reporting and frequency analysis was used to compare the perspectives of students on their performance. In addition, the questionnaire also explored how they conceive the idea of an integrated, holistic reporting and how it generates their motivation to improve. The iCGPA group was found to be receptive to what they have achieved throughout their study period. They agreed that the achievement level generated from their spider web allows them to develop intervention and enhance the programme learning outcomes before they graduate.

Keywords: learning outcomes attainment, iCGPA, programme learning outcomes, spider web, iCGPA reporting skills

Procedia PDF Downloads 208

Authors: Guanghua Zhang, Fubao Wang, Weijun Duan

Abstract:

Convolution neural network (CNN) has attracted more and more attention on recent years. Especially in the field of computer vision and image classification. However, unsupervised learning with CNN has received less attention than supervised learning. In this work, we use a new powerful tool which is deep convolutional generative adversarial networks (DCGANs) to generate images from Sloan Digital Sky Survey. Training by various star and galaxy images, it shows that both the generator and the discriminator are good for unsupervised learning. In this paper, we also took several experiments to choose the best value for hyper-parameters and which could help to stabilize the training process and promise a good quality of the output.

Keywords: convolution neural network, discriminator, generator, unsupervised learning

Procedia PDF Downloads 268

Authors: Jun Ming Moey, Zhiyaun Chen, David Nicholson

Abstract:

Bad actors are often hard to detect in data that imprints their behaviour patterns because they are comparatively rare events embedded in non-bad actor data. An unsupervised machine learning framework is applied here to detect bad actors in financial crime datasets that record millions of transactions undertaken by hundreds of actors (<0.01% bad). Specifically, the framework combines ‘shallow’ (PCA, Isolation Forest) and ‘deep’ (Autoencoder) methods to detect outlier patterns. Detection performance analysis for both the individual methods and their combination is reported.

Keywords: detection, machine learning, deep learning, unsupervised, outlier analysis, data science, fraud, financial crime

Procedia PDF Downloads 95

Authors: Kumiko Inagaki

Abstract:

In recent, years, Japanese universities have begun to face a dilemma: more than half of all high school graduates go on to attend an institution of higher learning, overwhelming Japanese universities accustomed to small student bodies. These universities have been forced to embrace qualitative changes to accommodate the increased number and diversity of students who enter their establishments, students who differ in their motivations for learning, their levels of eagerness to learn, and their perspectives on the future. One of these changes is an increase in awareness among Japanese educators of the importance of active learning, which deepens students’ understanding of course material through a range of activities, including writing, speaking, thinking, and presenting, in addition to conventional “passive learning” methods such as listening to a one-way lecture. 　The purpose of this study is to examine the effectiveness of the teaching method adapted to improve active learning. A teaching method designed to promote active learning was implemented in a social science course at one of the most popular universities in Japan. A questionnaire using a five-point response format was given to students in 2,305 courses throughout the university to evaluate the effectiveness of the method based on the following measures: ① the ratio of students who were motivated to attend the classes, ② the rate at which students learned new information, and ③ the teaching method adopted in the classes. The results of this study show that the percentage of students who attended the active learning course eagerly, and the rate of new knowledge acquired through the course, both exceeded the average for the university, the department, and the subject area of social science. In addition, there are strong correlations between teaching method and student motivation and between teaching method and knowledge acquisition rate. These results indicate that the active learning teaching method was effectively implemented and that it may improve student eagerness to attend class and motivation to learn.

Keywords: active learning, Japanese university, teaching method, university education

Procedia PDF Downloads 195

Authors: Rubén Valcarce-Álvarez, Francisco Jáñez-Martino, Rocío Alaiz-Rodríguez

Abstract:

Cybersecurity challenges, including scams, drug sales, the distribution of child sexual abuse material, fake news, and hate speech on both the surface and deep web, have significantly increased over the past decade. Users who post such content often employ strategies to evade detection by automated filters. Among these tactics, text obfuscation plays an essential role in deceiving detection systems. This approach involves modifying words to make them more difficult for automated systems to interpret while remaining sufficiently readable for human users. In this work, we aim at spotting obfuscated words and the employed techniques, such as leetspeak, word inversion, punctuation changes, and mixed techniques. We benchmark Named Entity Recognition (NER) using models from the BERT family as well as two large language models (LLMs), Llama and Mistral, on XX_NER_WordCamouflage dataset. Our experiments evaluate these models by comparing their precision, recall, F1 scores, and accuracy, both overall and for each individual class.

Keywords: natural language processing (NLP), text obfuscation, named entity recognition (NER), deep learning

Procedia PDF Downloads 3

Authors: Abrar Hussain Qureshi

Abstract:

Vocabulary has been highlighted as a key indicator in any foreign language learning program, especially English as a foreign language (EFL). It is often considered a potential tool in foreign language curriculum, and its deficiency impedes successful communication in the target language. The knowledge of the lexicon is very significant in getting communicative competence and performance. Nouns constitute a considerable bulk of English vocabulary. Rather, they are the bones of the English language and are the main semantic carrier in spoken and written discourse. As nouns dominate the bulk of the English lexicon, their role becomes all the more potential. The undertaken research is a systematic effort in this regard to work out a list of highly frequent list of Pakistani English nouns for the EFL learners at the secondary level. It will encourage autonomy for the EFL learners as well as will save their time. The corpus used for the research has been developed locally from leading English newspapers of Pakistan. Wordsmith Tools has been used to process the research data and to retrieve word list of frequent Pakistani English nouns. The retrieved list of core Pakistani English nouns is supposed to be useful for English language learners at the secondary level as it covers a wide range of speech events.

Keywords: corpus, EFL, frequency list, nouns

Procedia PDF Downloads 103

Authors: Erhan Eroğlu

Abstract:

This paper presents a new method called Mentor and Mentee Based Learning. This new method is becoming more and more common especially at workplaces. This study is significant as it clearly underlines how it works well. Education has always aimed at equipping people with the necessary knowledge and information. For many decades it went on teachers’ talk and chalk methods. In the second half of the nineteenth century educators felt the need for some changes in delivery systems. Some new terms like self- discovery, learner engagement, student centered learning, hands on learning have become more and more popular for such a long time. However, some educators believe that there is much room for better learning methods in many fields as they think the learners still cannot fulfill their potential capacities. Thus, new systems and methods are still being developed and applied at education centers and work places. One of the latest methods is assigning some mentors for the newly recruited employees and training them within a mentor and mentee program which allows both parties to see their strengths and weaknesses and the areas which can be improved. This paper aims at finding out the perceptions of the mentors and mentees on the programs they are offered at their workplaces and suggests some betterment alternatives. The study has been conducted via a qualitative method whereby some interviews have been done with both mentors and mentees separately and together. Results show that it is a great way to train inexperienced one and also to refresh the older ones. Some points to be improved have also been underlined. The paper shows that education is not a one way path to follow.

Keywords: learning, mentor, mentee, training

Procedia PDF Downloads 228

Authors: Felipe A. Guth, Shane Ward, Kevin McDonnell

Abstract:

In the precision agriculture area, one of the main tasks is the automated detection of diseases in crops. Machine Learning algorithms have been studied in recent decades for such tasks in view of their potential for improving economic outcomes that automated disease detection may attain over crop fields. The latest generation of deep learning convolution neural networks has presented significant results in the area of image classification. In this way, this work has tested the implementation of an architecture of deep learning convolution neural network for the detection of diseases in different types of crops. A data augmentation strategy was used to meet the requirements of the algorithm implemented with a deep learning framework. Two test scenarios were deployed. The first scenario implemented a neural network under images extracted from a controlled environment while the second one took images both from the field and the controlled environment. The results evaluated the generalisation capacity of the neural networks in relation to the two types of images presented. Results yielded a general classification accuracy of 59% in scenario 1 and 96% in scenario 2.

Keywords: convolutional neural networks, deep learning, disease detection, precision agriculture

Procedia PDF Downloads 259

Authors: Manuela Nayantara Jeyaraj

Abstract:

Having moved into the 21^st century, it is way past being arguable that innovative technology needs to be incorporated into conventional classroom teaching. Though the Western world has found presumable success in achieving this, it is still a concept under battle in developing countries such as Sri Lanka. Reaching the acme of implementing interactive virtual learning within classrooms is a struggling idealistic fascination within the island. In order to overcome this problem, this study is set to reveal facts that limit the implementation of virtual, interactive learning within the school classrooms and provide hacks that could prove the augmented use of the Virtual World to enhance teaching and learning experiences. As each classroom moves along with the usage of technology to fulfill its functionalities, a few intense hacks provided will build the administrative onuses on a virtual system. These hacks may divulge barriers based on social conventions, financial boundaries, digital literacy, intellectual capacity of the staff, and highlight the impediments in introducing students to an interactive virtual learning environment and thereby provide the necessary actions or changes to be made to succeed and march along in creating an intellectual society built on virtual learning and lifestyle. This digital learning environment will be composed of multimedia presentations, trivia and pop quizzes conducted on a GUI, assessments conducted via a virtual system, records maintained on a database, etc. The ultimate objective of this study could enhance every child's basic learning environment; hence, diminishing the digital divide that exists in certain communities.

Keywords: digital divide, digital learning, digitization, Sri Lanka, teaching methodologies

Procedia PDF Downloads 355

Authors: Bandhan Dey, Muhsina Bintoon Yiasha, Gulam Sulaman Choudhury

Abstract:

Chest disease is one of the most problematic ailments in our regular life. There are many known chest diseases out there. Diagnosing them correctly plays a vital role in the process of treatment. There are many methods available explicitly developed for different chest diseases. But the most common approach for diagnosing these diseases is through X-ray. In this paper, we proposed a multi-classification deep learning model for diagnosing COVID-19, lung cancer, pneumonia, tuberculosis, and atelectasis from chest X-rays. In the present work, we used the transfer learning method for better accuracy and fast training phase. The performance of three architectures is considered: InceptionV3, VGG-16, and VGG-19. We evaluated these deep learning architectures using public digital chest x-ray datasets with six classes (i.e., COVID-19, lung cancer, pneumonia, tuberculosis, atelectasis, and normal). The experiments are conducted on six-classification, and we found that VGG16 outperforms other proposed models with an accuracy of 95%.

Keywords: deep learning, image classification, X-ray images, Tensorflow, Keras, chest diseases, convolutional neural networks, multi-classification

Procedia PDF Downloads 92

Authors: Liu Jia’nan

Abstract:

The default accounts hold the view that there exists a kind of scalar implicature which can be processed without context and own a psychological privilege over other scalar implicatures which depend on context. In contrast, the Relevance Theorist regards context as a must because all the scalar implicatures have to meet the need of relevance in discourse. However, in Katsos, the experimental results showed: Although quantitatively the adults rejected under-informative utterance with lexical scales (context-independent) and the ad hoc scales (context-dependent) at almost the same rate, adults still regarded the violation of utterance with lexical scales much more severe than with ad hoc scales. Neither default account nor Relevance Theory can fully explain this result. Thus, there are two questionable points to this result: (1) Is it possible that the strange discrepancy is due to other factors instead of the generation of scalar implicature? (2) Are the ad hoc scales truly formed under the possible influence from mental context? Do the participants generate scalar implicatures with ad hoc scales instead of just comparing semantic difference among target objects in the under- informative utterance? In my Experiment 1, the question (1) will be answered by repetition of Experiment 1 by Katsos. Test materials will be showed by PowerPoint in the form of pictures, and each procedure will be done under the guidance of a tester in a quiet room. Our Experiment 2 is intended to answer question (2). The test material of picture will be transformed into the literal words in DMDX and the target sentence will be showed word-by-word to participants in the soundproof room in our lab. Reading time of target parts, i.e. words containing scalar implicatures, will be recorded. We presume that in the group with lexical scale, standardized pragmatically mental context would help generate scalar implicature once the scalar word occurs, which will make the participants hope the upcoming words to be informative. Thus if the new input after scalar word is under-informative, more time will be cost for the extra semantic processing. However, in the group with ad hoc scale, scalar implicature may hardly be generated without the support from fixed mental context of scale. Thus, whether the new input is informative or not does not matter at all, and the reading time of target parts will be the same in informative and under-informative utterances. People’s mind may be a dynamic system, in which lots of factors would co-occur. If Katsos’ experimental result is reliable, will it shed light on the interplay of default accounts and context factors in scalar implicature processing? We might be able to assume, based on our experiments, that one single dominant processing paradigm may not be plausible. Furthermore, in the processing of scalar implicature, the semantic interpretation and the pragmatic interpretation may be made in a dynamic interplay in the mind. As to the lexical scale, the pragmatic reading may prevail over the semantic reading because of its greater exposure in daily language use, which may also lead the possible default or standardized paradigm override the role of context. However, those objects in ad hoc scale are not usually treated as scalar membership in mental context, and thus lexical-semantic association of the objects may prevent their pragmatic reading from generating scalar implicature. Only when the sufficient contextual factors are highlighted, can the pragmatic reading get privilege and generate scalar implicature.

Keywords: scalar implicature, ad hoc scale, dynamic interplay, default account, Mandarin Chinese processing

Procedia PDF Downloads 323

Authors: Liang Wang, Beste F. Yuksel, David Guy Brizan

Abstract:

A computational model of affect that can distinguish between spontaneous and posed smiles with no errors on a large, popular data set using deep learning techniques is presented in this paper. A Long Short-Term Memory (LSTM) classifier, a type of Recurrent Neural Network, is utilized and compared to human classification. Results showed that while human classification (mean of 0.7133) was above chance, the LSTM model was more accurate than human classification and other comparable state-of-the-art systems. Additionally, a high accuracy rate was maintained with small amounts of training videos (70 instances). The derivation of important features to further understand the success of our computational model were analyzed, and it was inferred that thousands of pairs of points within the eyes and mouth are important throughout all time segments in a smile. This suggests that distinguishing between a posed and spontaneous smile is a complex task, one which may account for the difficulty and lower accuracy of human classification compared to machine learning models.

Keywords: affective computing, affect detection, computer vision, deep learning, human-computer interaction, machine learning, posed smile detection, spontaneous smile detection

Procedia PDF Downloads 125

Authors: Jaeseo Lim, Jooyong Park

Abstract:

Although there is ample evidence that student involvement enhances learning, college education is still mainly centered on lectures. However, in recent years, the effectiveness of discussions and the use of collective intelligence have attracted considerable attention. This study intends to examine the empirical effects of discussions on learning outcomes in various conditions. Eighty eight college students participated in the study and were randomly assigned to three groups. Group 1 was told to review material after a lecture, as in a traditional lecture-centered class. Students were given time to review the material for themselves after watching the lecture in a video clip. Group 2 participated in a discussion in groups of three or four after watching the lecture. Group 3 participated in a discussion after studying on their own. Unlike the previous two groups, students in Group 3 did not watch the lecture. The participants in the three groups were tested after studying. The test questions consisted of memorization problems, comprehension problems, and application problems. The results showed that the groups where students participated in discussions had significantly higher test scores. Moreover, the group where students studied on their own did better than that where students watched a lecture. Thus discussions are shown to be effective for enhancing learning. In particular, discussions seem to play a role in preparing students to solve application problems. This is a preliminary study and other age groups and various academic subjects need to be examined in order to generalize these findings. We also plan to investigate what kind of support is needed to facilitate discussions.

Keywords: discussions, education, learning, lecture, test

Procedia PDF Downloads 176

Authors: Boumaraf Malak

Abstract:

The development of intelligent transportation systems (ITS) and artificial intelligence (AI) are spurring us to pave the way for the widespread adoption of autonomous vehicles (AVs). This is open again opportunities for smart roads, smart traffic safety, and mobility comfort. A highly intelligent decision-making system is essential for autonomous driving around dense, dynamic objects. It must be able to handle complex road geometry and topology, as well as complex multiagent interactions, and closely follow higher-level commands such as routing information. Autonomous vehicles have become a very hot research topic in recent years due to their significant ability to reduce traffic accidents and personal injuries. Using new artificial intelligence-based technologies handles important functions in scene understanding, motion planning, decision making, vehicle control, social behavior, and communication for AV. This paper focuses only on deep reinforcement learning-based methods; it does not include traditional (flat) planar techniques, which have been the subject of extensive research in the past because reinforcement learning (RL) has become a powerful learning framework now capable of learning complex policies in high dimensional environments. The DRL algorithm used so far found solutions to the four main problems of autonomous driving; in our paper, we highlight the challenges and point to possible future research directions.

Keywords: deep reinforcement learning, autonomous driving, deep deterministic policy gradient, deep Q-learning

Procedia PDF Downloads 85

Authors: Imran Ullah Imran

Abstract:

Language contact is a familiar concept in the present global world. Across the globe, languages get mixed up at different levels. Borrowing, code-switching are some of the means through which languages interact. This study examines Pashto-English contact at word and syllable levels. By recording the speech of 30 Pashto native speakers, selected via 'social network' sampling, the study located a number of Pashto-English compound words, which is a unique contact of its kind. In data analysis, tokens were categorized on the basis of their pattern and morphological structure. The study shows that Pashto-English Bilingual Compound words (BCWs) are very prevalent in the Pashto language. The study also found that the BCWs in Pashto are completely productive and have their own meanings. It also shows that the dominant pattern of hybrid words in Pashto is the conjugation of an independent English root word followed by a Pashto inflectional morpheme, which contributes to the core semantic content of the construction. The BCWs construction shows that how both the languages are closer to each other. Pashto-English contact results into bilingual compound and hybrid words, which forms a considerable number of tokens in the present-day spoken Pashto. On the basis of these findings, the study assumes that the same phenomenon may increase with the passage of time that would, in turn, result in the formation of more bilingual compound or hybrid words.

Keywords: code-mixing, bilingual compound words, pashto-english contact, hybrid words, inflectional lexical morpheme

Procedia PDF Downloads 249

Authors: Michael Stewart

Abstract:

Mutation testing is a type of software testing proposed in the 1970s where program statements are deliberately changed to introduce simple errors so that test cases can be validated to determine if they can detect the errors. Test cases are executed against the mutant code to determine if one fails, detects the error and ensures the program is correct. One major issue with this type of testing was it became intensive computationally to generate and test all possible mutations for complex programs. This paper used reinforcement learning and parallel processing within the context of mutation testing for the selection of mutation operators and test cases that reduced the computational cost of testing and improved test suite effectiveness. Experiments were conducted using sample programs to determine how well the reinforcement learning-based algorithm performed with one live mutation, multiple live mutations and no live mutations. The experiments, measured by mutation score, were used to update the algorithm and improved accuracy for predictions. The performance was then evaluated on multiple processor computers. With reinforcement learning, the mutation operators utilized were reduced by 50 – 100%.

Keywords: automated-testing, machine learning, mutation testing, parallel processing, reinforcement learning, software engineering, software testing

Procedia PDF Downloads 198

Authors: Lise P. Labéjof, Krisnayne S. Ribeiro, Nicolle P. dos Santos

Abstract:

An unrecorded experiment of use of the smartphone as a tool for practical classes of histology is presented in this article. Behavior, learning of the students of three science courses at the University were analyzed and compared as well as the mode of teaching of this discipline and the appreciation of the students, using either digital photographs taken by phone or drawings for record microscopic observations, analyze and interpret histological sections of human or animal tissues.

Keywords: cell phone, digital micrographies, learning of sciences, teaching practices

Procedia PDF Downloads 596