CONTENTS
|
Spoken Digit Recognition using the k-Nearest-Neighbor method and
Mel Frequency Cepstral Coefficients Sorin MURARU, Cătălina-Lucia COCIANU 5 This study investigates the utilization of the k-nearest-neighbor algorithm within the framework of machine learning for speech recognition applications. The AudioMNIST dataset is used for performing the evaluations in which the model predicts the spoken digit, namely from 0 to 9. Two different training-to-test percentage splits of the dataset are used, 70%-30% and 80%-20%, while the k parameter ranges from 1 to 12. To better adapt the predic-tion model, the Mel-frequency cepstrum coefficients are extracted from each audio sample, and the 13 filters are averaged over 25 ms frame windows with 10 ms frame overlap. In both training-to-test configurations the value for the k parameter that obtained the highest accu-racy (> 95%) is k=5, while the easiest to predict digits was “7”. These findings underscore the efficacy of k-nearest-neighbor in speech recognition tasks and highlight the importance of parameter selection and feature extraction techniques in optimizing model performance. Further exploration of kNN's applicability in diverse speech recognition contexts holds promise for advancing the field's understanding and practical implementations. Keywords: k-nearest neighbor, Machine learning, MFCC, Speech recognition, Natural language processing Predicting Alzheimer’s Disease Using Deep Learning Artificial Intelligence Together with a Pre-Trained VGG19 and Inception_v3 Models Paul Gabriel TEODORESCU, Silvia OVREIU, Mădălina ZAMFIR, Cristian ȚÎRLEA 17 This paper presents two experiments in which, using artificial intelligence (specifically Deep Learning with convolutional neural networks), we were able to predict Alzheimer's disease based on MRI images. In order to have better results and to minimize the computational effort in the laboratory, two pre-trained AI models were used, models trained previ-ously on more than a million images from the ImageNet database (which provide tens of mil-lions of clean, labelled and sorted images). The top-layers of the models were trained, for our specific task of Alzheimer’s prediction, with 500 public MRI images from Kaggle, an online community of data scientists and machine learning engineers and a subsidiary of Google. In this paper we describe the code used in the laboratory for the specific task. Keywords: MRI images, ImageNet, Feature extractor, Demented, Accuracy, Pre-trained, Target, Convolutional, Matrix A Survey of IoT Frameworks for Low-Powered Devices Andrei-Robert CAZACU 35 Thanks to technological advancements, our lives are getting more intertwined with the connected world as more smart devices are coming to market. As such, the clear separation of devices as things (end-devices in IoT systems) and human operated devices is getting increasingly buried. This led to the creation of the term Internet of Everything (IoE) which is defined by Cisco as the “the networked connection of people, process, data, and things” [1]. The main difference between IoT and IoE is inclusion of people in the ecosystem, which greatly increases the number of connected parties. This increase in connected parties creates a strain on our existing infrastructure which is relying on cloud computing for performing most operations. Even though this resource provides heaps of computational power, the weak link in this scenario is the network, where all the connected devices can easily overload the available bandwidth, leading to slow response speeds and low general availability. The answer to this problem lies with technology that already exists and is not yet fully exploited as a distributed computing powerhouse, IoT. This paper aims to summarise the concept of computing at the edge, common architectural patterns, existing solutions, while also discussing real-world applications. Keywords: Edge Computing, Cloud Computing, EdgeX, Internet of Things, Internet of Everything, IoT, IoE The Cyber Competences Act - a Vital EU Regulation Concerning Mandatory Certification of Critical Network and Information Systems’ Operators across the European Union Dănuț MAFTEI 45 Cybersecurity is vitally important to all critical network and information systems (Critical N&IS). Effective information security comprises multiple layers of defense working together to protect Critical N&IS. When developing a proper cybersecurity, the only considered are especially technical layers, but effective penetration attacks often involve a mix of both the social and technical vectors of attack. The fact that the human error is one of the vital layers that could be a potential weakness for any system that incorporates humans needs a special attention and should be handled at the EU level. Given the increasing number of incidents created by insufficiently trained operators, a strategic necessity arises: the mandatory training and certification of Critical N&IS’ operators across the EU. This objective could be fulfilled through a new Regulation (The Cyber Competences Act) for laying down measures to complement achieving a high common level of cybersecurity within the European Union. Keywords: Cyber competences, Human error, Vulnerabilities, Critical network and information systems, Certification Blockchain Based DApps for Education Silviu OJOG, Paul POCATILU, Felician ALECU 61 Blockchain technology has captured the attention of various industries due to its potential to revolutionize traditional systems through decentralization, transparency and immutability. This paper examines the emerging trend of integrating blockchain-based decentralized applications (DApps) into the education sector. With blockchain's distributed ledger system, educational institutions can achieve safe, secure and transparent management of student achievements, certifications and credentials. The paper presents the architecture of a system that enables students to build verifiable digital portfolios of educational achievements stored and shared securely using smart contracts and digital tokens. By harnessing the benefits and addressing the challenges of blockchain DAppps in education, we can pave the way for a truly transformative era in the learning experience of tomorrow. Keywords: Blockchain, Smart Contract, Education, Security, Ethereum, Exploit, Immutability, Solidity The Mobile and Online Learning Impact in the Ukraine War Rareș-Constantin CIOBANU 72 In this paper it is displayed the resilience of the Ukrainian educators and students during the ongoing conflict with Russia, with a particular focus on the role of online learning platforms. Despite the challenges posed by air raids, evacuations, and infrastructure damage, educators have swiftly transitioned to virtual classrooms using platforms like Zoom, Google Meet, All-Ukrainian Online School and others. These platforms not only provide continuity in education but also serve as spaces for emotional support, fostering discussions about the conflict and offering solidarity to students amidst uncertainty. While significant challenges persist, including reaching students in heavily affected regions, the adaptability and dedication of Ukrainian educators underscore the resilience of the human spirit in adversity. Online learning continues to serve as a beacon of hope, ensuring that the pursuit of knowledge endures even in the darkest of times. Keywords: Education system, war, Ukraine, Online learning, M-learning Publishing Guide for Authors 82 INFOREC Association 84 |
