World Congress on Electrical Engineering and Computer Systems and Science最新文献

{"title":"Compact Right-Angled Coplanar Waveguide Using Central ShortCircuited Slotline","authors":"Chieh-Yu Liao, Chun-Long Wang","doi":"10.11159/eee22.116","DOIUrl":"https://doi.org/10.11159/eee22.116","url":null,"abstract":"- In this paper, a compact right-angled CPW using the central short-circuited slotline is proposed. The short-circuited slotline is fabricated on the central strip of the CPW so that it will not consume any additional area. As compared with the right-angled CPW using the short-circuited slotline [14], the reflection coefficient is maintained at a low value around -10 dB while the transmission coefficient is increased from -4.49 dB to -3.28 dB. Besides, the power loss is substantially reduced from 0.55 to 0.37.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"23 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120840473","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Archival Handwritten Digits Identification Through Deep Learning Models Johnson","authors":"Nathan LeBlanc, I. Valova","doi":"10.11159/cist22.131","DOIUrl":"https://doi.org/10.11159/cist22.131","url":null,"abstract":"- This research is inspired by the work of climate scientists to analyse archival handwritten documents and predict reliably changes in the climate. The aim of this work is to better understand recognition of handwritten documents especially focusing on archival maritime logs. Indeed, OCR (Optical Character Recognition) has existed for many years. The shortcomings of this widely used method, however, are manifesting in frequent confusion of digits and letters when it comes to archival handwritten documents. The conversation of perfecting the methods of automated recognition of handwritten characters has been evolving in recent years. With the advent of deep learning methods, new tools are considered within the problem space. In this extension of thesis work, two such methods are put to the test - convolutional and long-short term memory (LSTM) neural networks (NN). The applicability of several state-of-the-art models is considered with detailed experiments and comparative analysis. A compound model of convolutional NN followed by LSTM is also considered. While all models register high accuracy, it is observed that the compound model performs faster with accuracy above the lone CNN.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"174 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131898967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Investigating the Interaction between Data and Algorithms","authors":"Daniel Pototzky, Azhar Sultan, L. Schmidt-Thieme","doi":"10.11159/mvml22.106","DOIUrl":"https://doi.org/10.11159/mvml22.106","url":null,"abstract":"– Research in computer vision is centered on algorithmic improvements, for example, by developing better models. Thereby, the data is considered fixed. This is in contrast to many real-world applications of computer vision systems in which algorithms and data co-evolve. To address this shortcoming of previous research, we study the properties of the data and their interaction with deep learning algorithms. Thereby, we investigate the size of the data, the share of mislabels, class imbalance and the presence of unlabeled data which can be leveraged using semi-supervised learning. In experiments on 100 classes from ImageNet, we show that a tiny network architecture outperforms a much more powerful one it if has access to only a little bit more data. Only if vast amounts of data are available so that adding even more images has little effect on performance, large architectures dominate smaller ones. If little data is provided, adding a few labeled images has a huge effect on accuracy. Once accuracy saturates, massive amounts of additional data are needed to achieve even small improvements. Furthermore, we find that mislabels severely reduce performance. To fix that, we propose a cost-efficient way of identifying mislabels which is especially beneficial if many images are already available. Conversely, if little data is available, labeling more images is more advantageous than cleaning existing annotations. In the case of imbalanced data, we illustrate that labeling more instances from rare classes has a much greater effect on performance than only increasing dataset size. Moreover, we show that leveraging unlabeled images by semi-supervised learning offers a consistent benefit even if the labeled subset contains significant label noise.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123125952","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real Time End To End System for Underwater Communication","authors":"C. M. Atalay, Murat Üçüncü","doi":"10.11159/eee22.114","DOIUrl":"https://doi.org/10.11159/eee22.114","url":null,"abstract":"- Today’s technology has begun to improve very fast and communication technology tries to adapt this positive change. Each new innovation of theoretical changes has been trying to be implemented practically. Although it is relatively easier for RF communication to adapt this change since there are many manufactured special ICs for special tasks, it is still a challenging task for underwater communication. It is very difficult to establish a reliable link for underwater communication due to the lack of receiver designs suitable for underwater communication in the literature. Besides, it takes significant amount of time to implement such algorithms on FPGAs with HDL languages. In this paper, several successful telecommunication techniques of both RF and Under Water (UWA) communication are combined to construct a real time operating end to end UWA system. The proposed system is prepared on MATLAB with System Generator blocks. Therefore, it can directly be implemented on Xilinx’s FPGAs.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129364237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimizing Business Sales and Improving User Experience by using Intelligent User Interface","authors":"S. Pednekar, Swati Chandn","doi":"10.11159/mhci22.112","DOIUrl":"https://doi.org/10.11159/mhci22.112","url":null,"abstract":"– This research explores the impact on the user experience when the users, that is, the people in business, are exposed to an improved version of an intelligent user interface of the review management software. Machine learning algorithms, such as Lexicon-based sentimental analysis and NRC Emotion recognition, are employed to assist the proposed review management software, Review Dock. To provide additional assistance, a Content-based Recommendation system is integrated. More than 17,000 Amazon reviews are used to generate the results. To improve the satisfaction level of the already created prototype, three iterations of usability testing were conducted on nine participants. The findings show that by following the Web Content Accessibility Guidelines (WCAG) standards, an average satisfaction score of 2.49 out of 5 on the first iteration is significantly improved to 4.9 on the last iteration. Furthermore, the polarity categorization is similar across most evaluations, which are accomplished on previously unseen data sets. However, the results also reveal that the designs will only perform well for a small-medium industry. This research attempts to fill the limitations in the literature with respect to user experience. Regardless of the tools offered, the issue for businesses in utilizing an available solution that diminishes the engaging experience remains unchanged. As a result, a new solution should solve the limits, which will directly affect the company's sales. The research question states what steps the review management software may take to reduce the overly convoluted user interface? Therefore, proposing a solution called Review Dock will provide a plethora of responses and entirely focus on customer happiness by providing a comprehensive overview of how to enhance a product's sales.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128824656","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Applying Deep Learning for Image Segmentation: A Survey","authors":"Md. Jamiul Alam Khan, Jing Ren, Hossam A. Gabbar","doi":"10.11159/mvml22.101","DOIUrl":"https://doi.org/10.11159/mvml22.101","url":null,"abstract":"- Image segmentation is one of the most important branches of image processing. But it comes with various challenges and problems to be solved. Researchers are always working on improving the accuracy, quality and performance of image segmentation techniques. As in modern days, deep learning being involved in almost all problem solving, it is being used in image segmentation too. In this paper, we discussed few image segmentation techniques developed using deep learning, some implementation of these techniques to applications. And lastly, we addressed some limitations, challenges and research scopes for future.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132715824","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Use of Electronic Seizure Diaries and Decision Trees to Predict Seizure Outcome for Patients with Epilepsy","authors":"Dominique L. Tanner, M. Privitera, M. Rao, I. Basu","doi":"10.11159/cist22.139","DOIUrl":"https://doi.org/10.11159/cist22.139","url":null,"abstract":"- Epilepsy is a neurological disorder that causes unpredictable recurrent seizures. Most people with epilepsy dwell in fear of having unpredictable seizures. In attempts to predict future seizure occurrences, investigators have used data from electronic seizure diaries and machine-learning methods, like decision trees. Using individual patient e-diary data, the purpose of this study is to build patient specific decision trees to 1) determine decision trees overall accuracy in predicting seizures and depicting seizure predictors that influence seizure outcome, and 2) identify seizure predictors that have the most influence on seizure outcome. Patients (n=64) were examined, and their e-diary data was used to build patient specific decision trees. Using a 5-point Likert scale, patients e-diaries entailed information on how they rated the probability of experiencing subsequent seizures and rated their mood, predictive symptoms, stress, and seizure counts. Since e-diaries were recorded in the morning and in the evening, seizures for each patient were assessed by half days. R Programming software was used to generate the decision trees and depict seizure predictors that had the most influence on patient’s seizure outcome. A confusion matrix was performed to obtain the decision trees performance accuracy. Patients were categorized into groups based on certain seizure predictors that they shared. The results showed that for decision trees overall accuracy in predicting seizures and depicting seizure predictors that influenced seizure outcome, 49% of decision trees had an accuracy of 100%; 37% of decision trees had an accuracy ranging between 90-99%; and 13% of decision trees had an accuracy of <90%. Additionally, the results showed that there were more seizure predictors that had influence on patient’s seizure outcome in the morning than in the evening. This work introduces non-invasive precision medicine, with intentions to develop more personalized and reliable health care treatments for people with epilepsy.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123006299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Recursive Hierarchy for Accelerator-Level Parallelism","authors":"M. Malita, G. Stefan","doi":"10.11159/cist22.123","DOIUrl":"https://doi.org/10.11159/cist22.123","url":null,"abstract":"- The emergence, under the pressure of the ASICs imposed by the corporate space, of the field of Accelerator-Level Parallelism (ALP) requires a theoretical analysis to avoid the slippages that have characterized the evolutions of the last decades in the field of parallel computing. Ad hoc solutions imposed under time-to-market pressure have distorted the evolution of the field of parallel computing. The opportunity offered by the ALP challenge must be used to make last minute corrections in the chaotic evolution of the development of the parallel computing domain. The solution we propose is an attempt to reconsider parallelism from a double perspective. A purely theoretical one based on a mathematical model, that of the partially recursive functions proposed by Stephan Kleene, and another that emerges under the pressure of the increasingly complex applications demanded by the IT market. Our proposal consists in the hierarchical recursive structuring of ALP starting from the abstract MapScanReduce model that we have already proposed for the parallel computing.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129324548","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Visual Task Classification using Classic Machine Learning and CNNs","authors":"Devangi Vilas Chinchankarame, Noha M. Elfiky, Nada Attar","doi":"10.11159/mhci22.110","DOIUrl":"https://doi.org/10.11159/mhci22.110","url":null,"abstract":"- Our eyes actively perform tasks including, but not limited to, searching, comparing, and counting. This includes tasks in front of a computer, whether it be trivial activities like reading email, or video gaming, or more serious activities like drone management, or flight simulation. Understanding what type of visual task is being performed is important to develop intelligent user interfaces. In this work, we investigated standard machine and deep learning methods to identify the task type using eye-tracking data - including both raw numerical data and the visual representations of the user gaze scan paths and pupil size. To this end, we experimented with computer vision algorithms such as Convolutional Neural Networks (CNNs) and compared the results to classic machine learning algorithms. We found that Machine learning-based methods performed with high accuracy classifying tasks that involve minimal visual search, while CNNs techniques do better in situations where visual search task is included.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128595257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Radial Basis Functions: Meshless Interpolation and Approximation Methods","authors":"V. Skala","doi":"10.11159/cist22.001","DOIUrl":"https://doi.org/10.11159/cist22.001","url":null,"abstract":"Interpolation and approximation methods are widely used in data processing. The majority of methods require the data domain tessellation, e.g. using Delaunay triangulation, which leads to severe computational complication for higher dimensions. There are also severe problems with smoothness of the final interpolation or approximation in general. On the other hand, the meshless (meshfree) methods are simple as they leads to a solution of linear systems of equations. Also, smoothness is their natural property. Even more, the meshless based method based on radial basis functions (RBFs) are nearly independent on the problem dimensionality. In this talk, the basic principles of the RBF interpolation and approximation methods will be introduced with relevant mathematical formulations. Several examples of use will be given, especially some selected experimental results with large and high dimensional datasets will be presented.","PeriodicalId":294100,"journal":{"name":"World Congress on Electrical Engineering and Computer Systems and Science","volume":"01 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127165059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}