volume-1 CSIT issue 2
RUNGTA INTERNATIONAL OF JOURNAL OF COMPUTER SCIENCE AND INFORMATION TECHNOLOGY
RUNGTA INTERNATIONAL OF JOURNAL OF COMPUTER SCIENCE AND INFORMATION TECHNOLOGY
• Volume 1 (Jan 2024- Dec 2024)
Issue 2
1. Leveraging Big Data to Analyze Omicron Variant’s Daily Case Progression
Huma Khan1,#, Aradhana Sahu2, Bhawna Janghel3, Shaziya Islam4, Shweta Bandhekar5, Tripti Sharma6Author Affiliations
123456 Department of Computer Science and Engineering, Rungta College of Engineering and
Technology, Bhilai CG 490024
ABSTRACT:The extent of big data processing is progressively expanding. The study of big data processing is now underway. In this work, we analyzed the dataset of daily reported cases of the Omicron form of COVID-19. The data is filtered using an optimization point. The optimization method was conducted using the ant colony optimization (ACO) mechanism. When encountering complex optimization problems, one possible strategy is to employ a technique called “ant colony optimization,” abbreviated as “ACO.” The dataset that has been filtered using ACO is subsequently utilized to train an artificial neural network. Artificial neural networks (ANNs) are directly influenced by the biological neural networks seen in the brain of animals. An artificial neural network (ANN) is constructed using a network of interconnected units or nodes called artificial neurons. These artificial neurons have a similar structure and function to the neurons found in the human brain. Neural networks in the fields of artificial intelligence, machine learning, and deep learning imitate the operations of the human brain, allowing computer program to recognize patterns and address typical issues. This improves the efficiency and precision of the prediction process.
Keywords: Big data processing, Data processing, ACO, ANN, COVID 19 Corresponding author’s email address: shireenfatima22466@gmail.com
2. Dual Role of AI in The Cybersecurity: Defender and Threat
Kamal Kumar Prasad 1 ,Deepa Gajendra 2, Soniya Jaiswal 3,Anshita Shrivastava4 , Bhumika Shrivastava 5, Ashish Sonwani 6Author Affiliations
123456 Department of Computer Science and Engineering, Rungta College of Engineering and Technology, Bhilai C.G. 490024
ABSTRACT:
Artificial Intelligence (AI) has become both a powerful ally and a formidable adversary in the cybersecurity domain. This paper explores the dual role of AI as both a tool for enhancing cyber defenses and a weapon for executing sophisticated cyberattacks. On one hand, AI strengthens security through real-time threat detection, automated incident response, behavior-based intrusion systems, and predictive analytics. On the other hand, malicious actors increasingly exploit AI for automated phishing, deepfake scams, vulnerability discovery, and the generation of polymorphic malware. The study presents a comprehensive analysis of current AI-driven cybersecurity tools, highlights case studies where AI-enabled attacks have caused real-world damage, and critically examines the ethical and strategic implications of AI’s dual use. By evaluating current trends, defense mechanisms, and regulatory responses, this paper advocates for a balanced approach that fosters innovation while mitigating AI’s abuse. The research concludes with strategic recommendations to build resilient cybersecurity infrastructures that can harness AI’s power without falling victim to its misuse.
Keywords: AI in cybersecurity, adversarial AI, deepfake attacks, AI defense systems, threat detection, AI dual-use, ethical AI, automated hacking
3. Deep Supervised Feature Learning for Automated Tumor Detection in MRI Scans
Madhuri Gupta1, #, Onika Parmar1 , Padmavati Shrivastava1, Roushan Kumar1Author Affiliations
Department of Computer Science (AI/AIML), Rungta College of Engineering & Technology, Bhilai, India-490024
ABSTRACT:
Deep learning has transformed medical image processing in recent years, making it possible to detect tumours automatically and accurately. A supervised deep learning approach for brain tumor classification and feature extraction from MRI data is presented in this work. Instead of requiring human feature engineering, a CNN-based architecture is trained to learn spatial features straight from MRI scans. When tested on the BraTS dataset, the suggested model shows excellent sensitivity and robustness with an accuracy of 94.2%, an F1-score of 91.6%, and an AUC of 0.95. Techniques including data augmentation, transfer learning, and fine-tuning are used to improve performance, which greatly increases the model’s generalization across various MRI modalities. The results demonstrate how well the model supports clinical judgment with little assistance from humans.
Keywords: Deep learning, Supervised learning, MRI, Tumor detection, Convolutional Neural Network
4. Automated Segmentation of Chest X-ray Images: Evaluating the Efficacy of Deep Learning Architecture
Kushangi Atrey1, Padmavati Shrivastava1, Rahul Godara1,#, Sakshi Sinha1, Rohan Kumar1, Sonal Raj1Author Affiliations
Department of CSE-AI & AIML, Rungta College of Engineering and Technology, Bhilai (C.G), India- 490024
ABSTRACT:
Precise and timely segmentation of COVID-19 affected areas in X-ray scans is essential for clinical management, disease burden estimation, and proper treatment of the disease. The need for automated and reliable methods is highlighted by the time-consuming manual approach performed by radiologists. Despite the promise shown by deep learning, particularly U-Net-based architectures, and accurately segmenting heterogeneous infection areas with often undefinable boundaries remains a challenge. This paper investigates a sophisticated deep-learning architecture that integrates residual learning ideas and attention mechanisms for automated COVID-19 infection segmentation. The study made use of the publicly available dataset on which various data augmentation techniques were applied during the training phase. Several performance metrics were evaluated for the segmented outputs including accuracy (95.79%), Dice Coefficient (83.29%), Jaccard Index (84.53%), and Area under ROC Curve (0.98). Furthermore, the Grad-CAM visualizations identified the model’s attention on clinically pertinent pathological areas. The results show that the proposed architecture efficiently and powerfully enables automatically segmenting COVID-19 infections from X-ray scans. This will help the radiologists with patient assessment and treatment by simplifying their clinical procedures.
Keywords: COVID-19; Chest X-ray; Deep Learning; Image Segmentation; residual network.
5. AI-Driven Strategies for Forecasting and Minimizing Transportation Carbon Emissions
Jyoti Upaddhyay1,#, Farhat Anjum2, Chetna Sahu3, Bindu Kashyap4Author Affiliations
1,2,3,4 Department of Computer Science and Application, GD Rungta College of Science and Technology, Bhilai, India-490009
ABSTRACT:
The fast expansion of urban transportation has considerably contributed to the rise in carbon emissions, creating serious environmental and public health risks. With an emphasis on improving sustainability in urban mobility systems, this study investigates the use of artificial intelligence (AI) in anticipating and lowering carbon footprints associated with transportation. In order to anticipate carbon output with high accuracy, the research integrates a variety of datasets, such as vehicle density, fuel consumption patterns, traffic flow, and emission coefficients, using sophisticated machine learning and deep learning models. By collecting data related to traffic volume, vehicle types, and fuel usage in Bhilai, Chhattisgarh, the research builds a basic AI model capable of forecasting emission levels based on real-world factors. The proposed framework demonstrates the transformative potential of AI in driving data-informed, low-carbon transportation systems. Initial results show that AI models can help estimate emission trends with reasonable accuracy, providing a foundation for further research and policy development in low-carbon transportation systems.
Keywords: Artificial Intelligence (AI),Carbon Emissions, Emission Prediction, Machine Learning, Sustainable Mobility