Turkish Journal of Computer and Mathematics Education (TURCOMAT)
Not a member yet
9094 research outputs found
Sort by
FEMININE SENSIBILITY IN CHITRA BANERJEE DIVAKARUNI’S SISTER OF MY HEART
No full textIn Chitra Banerjee Divakaruni’s novel Sister of My Heart, the exploration of feminine sensibility through the lives of Anju and Sudha unveils a compelling narrative of resilience, struggle, and self-discovery in the face of societal pressures and male dominance in India. As the protagonists navigate the complexities of patriarchal society, their experiences shed light on the psychological impacts of entrenched gender norms and the challenges of carving out individual identities within a rigidly defined feminine role. Divakaruni skillfully contrasts Indian and American notions of feminism, delving into how cultural and sociological differences shape feminine sensibility and influence familial and marital relationships – underscoring the emotional toll wrought by societal, cultural, and psychological pressures on women. Through an examination of Anju and Sudha’s journeys, this research paper aims to delve deeper into the multifaceted nature of feminine roles and identities, offering insights into the intricate interplay between personal agency and external constraints in shaping women’s lived experiences
Eye Deep-Net: a deep neural network-based multi-class retinal disease diagnostic
No full textOphthalmologists rely heavily on retinal pictures to diagnose a wide range of eye conditions. Numerous retinal disorders may lead to microvascular alterations in the retina, and a number of studies have been conducted on the early identification of medical pictures to enable prompt and appropriate treatment. In order to identify various eye illnesses using color fundus pictures, this study develops a non-invasive, automated deep learning system. A multiclass ocular illness A productive diagnostic approach was created using the Remind dataset. A variety of augmentation strategies were used to make the structure robust in real-time after multi-class fundus pictures were collected from a multi-label dataset. Low computational demand images were processed in accordance with the network. The fundamental convolutional neural network (CNN) extracts appropriate characteristics from the input color fundus image dataset, and then processed characteristics were employed to make predictive diagnoses. This multi-layer neural network, called Eye Deep-Net, has been developed for training and evaluating images for the recognition of various eye problems. The performance of the suggested model is determined to be much better than numerous baseline state-of-the-art models. The strength from the Eye Deep-Net is assessed using different statistical metrics. The suggested methodology\u27s effectiveness in classifying and identifying diseases using digital fundus pictures is shown by a thorough comparison with the most modern techniques
Using CNN, GRU, and B/idirectional Multiscale Convolutional Neural Networks for Human Behavior Recognition
No full textThe main challenge in recognizing human behavior is constructing a network for the extraction and categorization of spatiotemporal features. In order to address the issue that the current channel attention mechanism simply aggregates each channel\u27s global average information while ignoring its specific spatial information, this work suggests two enhanced channel attention modules: the depth separable convolutions section and the time-space (ST) interaction section of matrices operation. These modules are also combined with research on the recognition of human behavior. Proposing a multiple habitats convolutional neural network technique for human behavior detection, it is combined with the excellent performance using convolutional neural network (CNN) for video and image processing. First, the behavior video is divided into segments. Next, low rank learning is applied to each segment to extract the associated low rank actions information. Finally, these minimal position behavior information are linked together in the time axis to get the low are behavior data for the entire video. This allows for the efficient extraction of behavior information from the video without the need for laborious extraction processes or assumptions. Neural networks can simulate human behavior in a variety of network topologies by transferring and reusing this capacity. To lessen the distinction between features derived from various network topologies, two efficient feature difference measurement methods are presented, taking into account the various properties of data features at various network levels. The suggested strategy has a decent categorization impact, according to experiments on a number of available datasets. The experimental findings demonstrate that the method\u27s accuracy in identifying human behavior is excellent. It has been shown that the suggested model increases recognition accuracy while simultaneously enhancing the compactness for the model structure and successfully lowering the computational cost of the output weights
IMAGE CAPTION GENERATOR USING CNN AND LSTM
No full textMachine learning is now all the rage in the AI world. We have recently used AI to construct very clever devices with exceptional performance. Deep learning is a subset of machine learning that produces very accurate findings, which in turn indicates very good performance. Apps for picture description make use of deep learning in our study. Providing a description of a picture\u27s content is what image description is all about. Object and action detection in the input picture is the foundation of the notion. When describing images, there are primarily two methods: bottom-up and top-down. Bottom-up methods create captions by combining the information of an input picture. Using different architectures, such as recurrent neural networks, top-down methods provide a semantic representation of an input picture, which is then translated into a caption. One potential advantage of picture description is that it might aid those with visual impairments in comprehending what is shown in online images. What follows is an explanation of the specifics. Looking at the image below, what can you make out
BLOCKCHAIN-BASED ARCHITECTURE AND FRAMEWORK FOR CYBERSECURE SMART CITIES
No full textA smart city is one that uses digital technologies and other means to improve the quality of life of its citizens and reduce the cost of municipal services. Smart cities primarily use IoT to collect and analyze data to interact directly with the city’s infrastructure and monitor city assets and community developments in real time to improve operational efficiency and proactively respond to potential problems and challenges. Today, cybersecurity is considered one of the main challenges facing smart cities. Over the past few years, the cybersecurity research community has devoted a great deal of attention to this challenge. Among the various technologies being considered to meet this challenge, Blockchain is emerging as a solution offering the data security and confidentiality essential for strengthening the security of smart cities. In this paper, we propose a comprehensive framework and architecture based on Blockchain, big data and artificial intelligence to improve smart cities cybersecurity. To illustrate the proposed framework in detail, we present simulation results accompanied by analyses and tests. These simulations were carried out on a smart grid dataset from the UCI Machine Learning Repository. The results convincingly demonstrate the potential and effectiveness of the proposed framework for addressing cybersecurity challenges in smart cities. These results reinforce the relevance and applicability of the framework in a real-world contex
REMAINING USEFUL LIFE PREDICTOR FOR EV BATTERIES USING MACHINE LEARNING
No full textElectric vehicles (EVs) are a key solution to combat rising carbon emissions and reduce dependence on fossil fuels. In India, the government has implemented policies such as the Faster Adoption and Manufacturing of Hybrid and Electric Vehicles (FAME) scheme to promote EV adoption Predicting the Remaining Useful Life (RUL) of EV batteries using machine learning ensures better battery health management and enhances operational efficiency. Applications include EV fleet management, battery recycling, and cost-effective maintenance. To develop a machine learning model that accurately predicts the Remaining Useful Life (RUL) of EV batteries to improve operational reliability, reduce maintenance costs, and support sustainable energy practices. Before the advent of machine learning, traditional methods for estimating EV battery life relied on rule-based approaches, where predefined thresholds such as voltage drops or charge cycles were used to predict battery health. Empirical models, often linear, were developed based on historical performance data but lacked the ability to adapt to dynamic usage patterns. Additionally, manual battery testing was a common practice to measure degradation, though it was time-consuming, labor-intensive, and often prone to inaccuracies in capturing the complex nature of battery aging. Traditional systems for predicting EV battery life are largely empirical and rely on static models that fail to capture dynamic battery behavior. These methods often lack precision, are labor-intensive, and provide limited adaptability to varying usage conditions, leading to inefficiencies in battery management. The increasing demand for EVs and their critical dependence on battery performance drives the need for accurate RUL prediction systems. Traditional methods are insufficient in addressing the complex, non-linear nature of battery degradation. The proposed machine learning-based system leverages real-time battery performance data, including metrics like voltage, current, temperature, and charge-discharge cycles, to train predictive models capable of estimating the Remaining Useful Life (RUL) of EV batteries. This approach significantly enhances accuracy by capturing complex patterns in battery degradation, enables real-time predictions for immediate insights, optimizes costs by minimizing unnecessary replacements and maximizing resource utilization, and promotes sustainability through efficient recycling and reduced battery waste
A ROBUST DETECTION OF CYBER INCIDENTS UTILIZING MACHINE LEARNING TECHNIQUES
No full textA reliable Cyber Attack Detection Model (CADM) is a system that works as safeguard for the users of modern technological devices and assistant for the operators of networks. The research paper aims to develop a CADM for analyzing the network data patterns to classify cyber-attacks. CADM finds out attack wise detection accuracy using ensemble classification method. LASSO has been used to extract important features. It can work with large datasets, and it has more visualization capability. Gradient Boosting and Random Forest algorithms have been used for classification of network traffic data to build an ensemble method. Gradient Boosting algorithm trains weak learning models and select the best decision trees to deliver more improved prediction accuracy and Random Forest algorithm trains each tree in parallel manner. In this research work, Jive datasets such as NSL-KDD, KDD Cup 99, UNSWNB15, URL 2016 and CICIDS 2017 are also applied to check the efficiency of the proposed model
Implementation of FIR Filters through Inner product Units and Parallel Accumulations
No full textFinite Impulse Response (FIR) filters are pivotal in digital signal processing, finding applications in diverse fields like audio processing, telecommunications, and biomedical signal analysis. This work presents an enhanced implementation methodology for FIR filters utilizing inner product computation and parallel accumulations. In the existing, FIR filters are typically implemented using convolution techniques, basic adders, and multipliers, which involve sequential processing and intensive computational resources. This method often leads to latency issues and limits real-time applications. Moreover, traditional implementations suffer from inefficiencies in utilizing hardware resources optimally, leading to suboptimal performance. The proposed methodology overcomes these limitations by leveraging inner product computations and parallel accumulation techniques. By exploiting inherent parallelism in the filtering process, the proposed method significantly reduces latency and enhances throughput
Performance of Hybrid Connected Network on Chip Router to Improve Latency and Throughput
No full textThe design of Network-on-Chip (NoC) routers plays a critical role in ensuring efficient data transmission. This project presents an innovative approach to designing NoC routers that prioritize area efficiency. It introduces a hybrid scheme tailored for NoCs, aiming to significantly reduce latency and power consumption. Existing NoC architectures typically employ either circuit switching or packet switching techniques, each with its own limitations. Circuit switching can lead to high latency due to setup time, while packet switching suffer from increased power consumption and congestion. To address these drawbacks, our proposed hybrid scheme combines virtual circuit switching with existing circuit and packet switching methods. By allowing multiple virtual circuit-switched (VCS) connections to share a common physical channel, our approach optimizes resource utilization and minimizes latency, Throughput. Furthermore, the integration of virtual circuit switching introduces dynamic routing flexibility, enhancing adaptability to varying traffic conditions. Hence, this work shows the superior performance and efficiency of our hybrid scheme compared to traditional NoC architecture
SPECTRUM SENSING USING COOPERATIVE MATCHED FILTER DETECTOR IN COGNITIVE RADIO
No full textThe vast rise in the number of internet-connected devices necessitates a more accessible spectrum. As a result, Cognitive Radio was already proposed as a solution to the problem of restricted spectrum resources by utilizing available spectrum which is assigned to primary users. This method allows the secondary user to utilize the spectrum whenever the primary user is not using it, and it does so without intruding with the primary user. Whenever the secondary user detects the spectrum, it faces many issues, such as complexity in sensing, leading to a lack of noise value, and the primary user is hidden to all secondary users. In order to tackle these challenges, many spectrum sensing frameworks were introduced in the literature. In this paper, an adaptive threshold matched filter detector and a cooperative matched filter detector frameworks are utilized to detect the spectrum and resolve the issues above. The probability of detection (Pd), probability of miss detection (Pm), and probability of false alarm (Pf) are the metrics used to assess sensing accuracy. To simulate suggested detectors results and proficiency, the MATLAB R2020a software was utilized. In comparison to earlier studies, the simulation conclusions reveal that the detection process starts with lower SNR values compared to previous work