KIET Journal of Computing and Information Sciences

Power Optimized Task Scheduling using Genetic Algorithm (POTS-GA) in Cloud Environment

Minhaj Ahmad Khan, Sana Saleem — Mon, 21 Oct 2024 00:00:00 -0600

In a cloud environment, the allocation of tasks has become pivotal on account of rapid growth of user requests. The processing of user requests leads to a significant execution time, and a huge amount of power is also consumed. Consequently, task scheduling for optimizing makes pan and power usage has become critical, particularly in a heterogeneous environment. This research work proposes Power-Optimized Task Scheduling using Genetic Algorithm (POTS-GA) that aims to minimize execution time and power consumption. The proposed strategy employs genetic algorithm to take scheduling decision while taking into consideration the execution time and overall power consumption of resources. The fitness computation considering both objectives and the customized genetic operators ensure to search for a better scheduling solution. The experiments performed on a large number of tasks and virtual machines show that the proposed POTS-GA approach outperforms other task scheduling strategies including Efficient Task Allocation using Genetic Algorithm (ETA-GA), Round Robin algorithm (RRA), and First Come First Serve (FCFS) and Greedy algorithm in terms of makes pan and power consumption.

Accurate Attack Detection in Intrusion Detection System for cyber threat intelligence feeds using machine learning techniques.

Ehtsham Irshad, Abdul Basit Siddiqui — Mon, 21 Oct 2024 00:00:00 -0600

With the advancement of modern technology, cyber-attacks are continuously rising. Malicious behavior in the network is discovered using security devices like intrusion detection systems (IDS), firewalls, and antimalware systems. To defend organizations, procedures for detecting threats more correctly and precisely must be defined. The proposed study investigates the significance of cyber-threat intelligence (CTI) feeds in accurate IDS detection. The NSL-KDD and CSE-CICIDS-2018 datasets were analyzed in this study. This research makes use of normalization, transformation, and feature selection algorithms. Machine learning (ML) techniques were employed to determine if the traffic was normal or an attack. With the proposed study the ability to identify network attacks has improved using machine learning algorithms. The proposed model provides 98% accuracy, 97% precision, and 96% recall respectively.

Predicting Student Performance Using Educational Data Mining: A Review

Mon, 21 Oct 2024 00:00:00 -0600

Educational Data Mining (EDM) strategies facilitate the efficient and in-depth analysis of student data. EDM provides useful insights into comprehending student learning patterns and identifying factors that influence academic success. This review aims to evaluate the efficacy of classification algorithms popularly explored in EDM for predicting student performance and identifying common trends in existing EDM research. The review follows a systematic approach, relevant research articles have been cited following an inclusion and exclusion criteria to ensure the selection of studies that specifically address the use of EDM techniques for predicting student academic achievement. According to the review findings, most researchers have utilized the features of cumulative grade point average, internal and external assessment, and demographic information to predict student performance. The most common techniques in EDM for predicting students’ performance are Naïve Bayes and Decision Trees. The review also focuses on the potential for bias, key examination of challenges, and possible future directions in the field. In the context of student performance prediction, ethical considerations regarding privacy, data handling, and the interpretation of results are also identified

A Hybrid Model for Human Behavior Recognition Using Emotions, Sentiments, and Mood Features

Hina Shakir, Asia Samreen, Syed Asif Ali, Muhammad Hussain — Mon, 21 Oct 2024 00:00:00 -0600

While social networking is a powerful communication tool, the obscured behavior of individuals on social networks remains a significant problem for users. Currently, research work is being focused on formulating mechanisms to determine the obscured behavior of users for secure and trustworthy social media. The proposed model employs mathematical formulation and multinomial classification of mood and emotions to analyze the conduct of an individual, thus enabling social trust on social media. First, natural language processing techniques are applied to predict the emotions, moods, and sentiments of an individual from the text, and then a mathematical model is applied to gather a comprehensive picture of one’s behavior using calculations at numerous instants. Finally, a subsequent trust state log is built in terms of positive and negative states which show the devotion in behavior in terms of mood, sentiments, and more significantly emotions. The efficiency of the proposed work has been demonstrated using simulation-based and real-world datasets along with individual behavior graphs for various conversations

Predicting and Characterizing piRNAs and their functions Using an Integrated Machine Learning Approach

Usman Inayat, Anam Umera, Sajid Mahmood — Mon, 21 Oct 2024 00:00:00 -0600

One class of short non-coding RNA molecule that is well recognized is called PIWI-interacting RNA (piRNA). PiRNAs are involved in the creation of novel medications as well as the identification of different kinds of tumors. Additionally, it is associated with stopping transposes, managing gene transcription, and maintaining genomic integrity. The important role that piRNAs play in biological processes has led to a growing body of research in bioinformatics on the discovery of piRNAs and their functionality. In this research, a powerful model is proposed to improve PiRNA prediction and functionality. The suggested model uses four classifiers (Logistic Regression, SVC, Random Forest, and Gradient Boosting Classifier) for classification. Moreover, TNC and DNC are used to acquire features. There are two layers involved in developing the suggested model. A sequence's potential to be piRNA is predicted in the first layer, and its potential to direct target mRNA deadenylation is predicted in the second. In the first layer, the model's accuracy is 98.59%, and in the second layer, it is 94.55%.