Turkish Journal of Computer and Mathematics Education (TURCOMAT)
Not a member yet
9094 research outputs found
Sort by
Memory Loss and Alzheimer\u27s Disease Progression Convolutional Neural Networks with Dropout Layers for Optimal Filtered Features in MRI Images of the Hippocampus for Slice Selection Based on Landmarks
No full textThe public health threat of Alzheimer\u27s disease (AD) is now widely accepted. When using machine learning techniques and MRI scanning to detect Alzheimer\u27s disease, the hippocampi are readily accessible and one among the most afflicted brain regions. AD classification by machine learning algorithms using complete MRI slices was unsatisfactory. This article describes how to choose MRI slices using hippocampus landmarks. This research aims to find the best accurate AD categorization MRI pictures. Next, utilizing Resnet50 or LeNet using various classifiers with the open-source and free ADNI dataset, the three views and categories were valued. The models used 4,500 Neuroimaging slices from three perspectives and categories for training. We found that AD classification was better with MRI scan segments than whole slices. The coronal view showed our method\u27s machine learning accuracy enhancement most clearly. This strategy greatly enhanced machine learning accuracy. The findings from a rotational perspective matched what clinicians use to identify AD. Additionally, LeNet models may classify AD effectively
A Deep Reinforcement-Based Anomaly Intrusion Detection for Enhancing Network Cybersecurity
No full textConventional protection methods, such as rules-based firewalls and signature-based detection, are not cutting it in today\u27s environment of increasingly sophisticated and frequent cyberattacks. Cyberattacks nowadays are extremely dynamic and complex, calling for cutting-edge solutions that can change and adapt as the threat does. DRL is an AI subfield that has been successfully addressing difficult decision-making challenges in several fields, including cybersecurity. Here, we make a step forward by using a DRL framework to model cyberattacks; by incorporating real-world events, we make the models more realistic and applicable. We provide a customized approach that greatly improves existing approaches by carefully tailoring DRL (deep reinforcement algorithms to the complex needs of cybersecurity situations, including adversarial training, dynamic environments, bespoke structure of reward and actions, and more.
In this study, we provide an anomaly detection method to detect attacks on network CPS using Deep Reinforcement Learning. Our proposed methodology was tested using several publicly available research datasets to ensure its efficacy
DENSITY BASED SMART TRAFFIC CONTROL SYSTEM USING CANNY EDGE DETECTION
No full textThe need for state-of-the-art equipment and technology to enhance traffic management is become more pressing as the problem of urban traffic congestion deteriorates. empirical evidence has shown that the traditional methods, such as timers and human control, are inadequate in effectively tackling this problem. The present study introduces a traffic control system that employs digital image processing and intelligent edge identification to enable real-time measurement of vehicle density. In contrast to earlier systems, this high-performance traffic control system offers a significant improvement in response time, automation, vehicle management, reliability, and overall efficiency. Furthermore, the whole process, including picture collection, edge recognition, and green signal allocation, is documented with suitable schematics and validated by hardware implementation using four illustrative images of different traffic situation
FINANCIAL FRAUD DETECTION USING VALUE AT RISK WITH MACHINE LEARNING IN SKEWED DATA
No full textThe significant losses that banks and other financial organizations suffered due to new bank account (NBA) fraud are alarming as the number of online banking service users increases. The inherent skewness and rarity of NBA fraud instances have been a major challenge to the machine learning (ML) models and happen when non-fraud instances outweigh the fraud instances, which leads the ML models to overlook and erroneously consider fraud as non-fraud instances. Such errors can erode the confidence and trust of customers. Existing studies consider fraud patterns instead of potential losses of NBA fraud risk features while addressing the skewness of fraud datasets. The detection of NBA fraud is proposed in this research within the context of value-at-risk as a risk measure that considers fraud instances as a worst-case scenario. Value-at-risk uses historical simulation to estimate potential losses of risk features and model them as a skewed tail distribution. The risk-return features obtained from value-at-risk were classified using ML on the bank account fraud (BAF) Dataset. The value-at-risk handles the fraud skewness using an adjustable threshold probability range to attach weight to the skewed NBA fraud instances. A novel detection rate (DT) metric that considers risk fraud features was used to measure the performance of the fraud detection model. An improved fraud detection model is achieved using a K-nearest neighbor with a true positive (TP) rate of 0.95 and a DT rate of 0.9406. Under an acceptable loss tolerance in the banking sector, value-at-risk presents an intelligent approach for establishing data-driven criteria for fraud risk management
Generative AI in Insurance: Automating Claims Documentation and Customer Communication
No full textWithin risk management strategies, genuine losses incurred from perils in primary perils are routine but dispropor- tionate, resulting in claims that impact organizational effort and cost. Expanding business lines beyond simply on-lending deposits is key to returning the enterprise to profitability. Reducing annual customer complaints and retentions show institutions that priori- tize customer service and have proactive Mitigants on their watch list are invariably rewarded by lower claims. Current and recent developments in the finance, technology and insurance sectors have combined to allow possible associations between these areas. New services offered by local Insurers, Market Shortfalls with Private Equity funds and Corporate Private Banks, Increasing Consumer Demand and the introduction of Weather Money to address Private Pine interventions have highlighted a trend of increasing risk severity. Such increasing risk severity has the potential to develop a property risk and make allowance for a worsening of insurance coverage conditions
S-METRIC SPACES FIXED POINT RESULTS ON ALTMAN INTEGRAL TYPE MAPPINGS
No full textIn this present article, we establish the concept o
STOCK SELECTION USING SEMI-VARIANCE AND BETA TO CONSTRUCT PORTFOLIO AND EFFECT MACRO-VARIABLE ON PORTFOLIO RETURN
No full textThis research has aims to construct portfolio by varying method and using semi-variance and Beta for selection stocks. This research found 28 stocks to become member portfolio. Equal Weighted, Market Capitalization Weighted, Markowitz Method and Elton Gruber is used to construct portfolio. This research found that the efficient frontier similar to Markowitz Method. Roy Criterion found the portfolio return varying from 2.2% to 9.65% but Kataoka Criterion found the portfolio return varying from 5.4% to 11.12%. This research found that Elton Gruber has the highest portfolio return compared to others portfolio. There is no difference of average return for four portfolios. Market return significant affect to all portfolio return but the interest rate significant affect portfolio returns for equal weighted portfolio and Elton Gruber Method
Unveiling Hidden Threats with ML-Powered User and Entity Behavior Analytics (UEBA)
No full textThe ever-growing cost of cybercrime has created the need for proactive solutions for organizations seeking to protect their digital assets. While traditional security systems struggle to detect anomalies buried within vast datasets, new solutions like User and Entity Behavior Analytics (UEBA) emerge as a game-changer. By leveraging the power of machine learning, UEBA analyzes diverse data sources like user logins, file accesses, event logs, business context, externalthreat intelligence, and network activity, to unveil hidden threats most traditional methods could miss. The ability to analyze multiple data sources enables UEBA solutions to effectively detect malicious insiders, compromised users, Advanced Persistent Threats (APTs), and zero-day attacks. By using various analytics techniques like supervised learning, unsupervised learning, and statistical modeling, UEBA solutions can detect subtle anomalies that deviate fromestablished behavior baselines. Despite the many benefits, UEBA solutions still have limitations like data quality concerns, high implementation costs, and the need for model maintenance. Integration with System Information and Event Management (SIEM) systems helps mitigate some of these challenges to further enhance UEBA\u27s capabilities and provide a unified platform for threat identification and response. This paper provides a detailed insight into the capabilities ofUEBA, its three pillars, significance, and limitations
Common Fixed Point Theorems Satisfying Contractive Type Conditions in Complex Valued Metric Spaces
No full textIn this paper we prove common fixed point theorems satisfying contractive conditions involving rational expressions and product for four mappings that satisfy property (E.A) along with weak compatibility of pairs are proved property are obtained in complex valued metric spaces which generalize various results of ordinary metric space
Exploring Factors Contributing to Indifference Towards Learning Mathematics Among Secondary School Students in Nepal
No full textMathematics is a compulsory subject at the school level in Nepal, deemed essential for everyday life and higher studies, particularly in the fields of science and technology. However, there is a noticeable apathy among students when it comes to learning mathematics. This qualitative research aims to identify the factors that contribute to this indifference towards learning mathematics. Data was collected through in-depth interviews with four participants from both public and private schools, all enrolled in the tenth grade. Analysis and interpretation of the data revealed several factors that lead to this indifference. These factors can be classified as student-related, school-related, and home and society-related. Student-related factors include mathematics anxiety, negative perceptions, insufficient effort, poor academic achievements, limited real-world applications, low self-efficacy, and perpetuation of misconceptions about mathematics. School-related factors encompass teaching practices, teacher qualifications, traditional methods focused on rote learning, impractical curriculum and courses, inadequate school administration, and subpar physical facilities. Home and society-related factors have a negative effect on mathematics engagement, such as unfavorable home environments, low socioeconomic status, and parental education. Together, these factors contribute to the observed indifference towards learning mathematics.
Keywords: Indifference, Qualitative, Mathematics, Factors, Home, Students, Schoo