Journal of Science Engineering Technology and Management Science

Accurately forecasting potential customers, reducing resource waste, and improving efficiency may all be necessary for telemarketing to be productive. To find clients, we use machine learning models such as Random Forest, XGBoost, AdaBoost, and Gradient Boosting. F1 score, recall, accuracy, and precision are used to assess these algorithms. To address class imbalance in telemarketing datasets, oversampling techniques like SMOTE, ADASYN, and Borderline SMOTE generate synthetic data for underrepresented classes. Model performance is improved by MinMax scaling data normalization and Mutual Information feature selection. Training and testing are conducted using telemarketing data from the Portuguese Bank Marketing Data Set. Modifying the XGBoost hyperparameters raises the model's accuracy to more than 98% for both the top 10 features and the complete set. To promote user involvement, the system has a real-time prediction interface built on Flask. Ensemble-based online machine learning, a successful telemarketing optimization technique, enables models to adjust to shifting customer preferences.

Keywords : Ensemble Learning; Online Machine Learning; Telemarketing; Class Imbalance; Synthetic Oversampling; SMOTE; ADASYN; Borderline SMOTE; Mutual Information; Feature Selection; Data Normalization; XGBoost; Random Forest; Gradient Boosting; Real-Time Prediction; Flask Interface; Bank Marketing Data

Author : SK. Himambasha1 , P. Sai Ram2

Title : Improved Ensemble-Based Online Machine Learning Approach for Telemarketing Success Prediction Using Hyperparameter-Tuned XGBoost

Volume/Issue : 2026;03(06)

Page No : 1138-1147

Elderly people and those who need constant attention are more vulnerable to falls. A real-time vision-based fall detection system utilizing deep learning and human position estimate methods is presented in this study. The suggested method combines MediaPipe Pose for skeletal landmark extraction from video frames with the YOLOv8 model for precise human detection. Using a rule-based decision procedure with time-based validation to minimize false alarms, fall events are detected by examining posture changes and body orientation. The device immediately produces audible and visual alarms when a fall is confirmed. A nonintrusive and effective solution for healthcare monitoring, the system is developed as a Flask-based web application that supports both live video streaming and recorded video analysis.

Keywords : Fall Detection, Computer Vision, Deep Learning, YOLOv8, Human Pose Estimation, MediaPipe, Real-Time Monitoring, Healthcare Applications, Vision-Based System, Non-Intrusive Monitoring.

Author : 1Mrs.M. Anusha,2Bhargavi Mohana Maddu,3Evuri Hari Krishna,4Chadalavada Rajkumar

Title : Intelligent Video-Based Fall Surveillance System for Elderly Safety

Volume/Issue : 2026;03(06)

Page No : 1148-1155

The College Event Planning System is a web-based application designed to automate and simplify the management of academic, cultural, and technical events in educational institutions. Traditional event management methods rely on manual coordination, notices, and paperwork, which are inefficient and prone to errors. The proposed system provides a centralized platform with role-based access for administrators, faculty members, and students. Administrators can create and manage events, faculty members can coordinate assigned events, and students can view and register for events online. The system ensures accurate data handling, improved communication, and transparent event management. Developed using HTML, CSS, JavaScript, Java (JSP and Servlets), and MySQL, the application reduces manual workload, improves efficiency, and enhances user experience. Overall, the system offers a reliable and userfriendly solution for effective college event planning and management.

Keywords : College Event Planning System, Web Application, Event Management, Role based access, Authentication and Data Security.

Author : 1Dr. P.V.S. Sarma,2Gandikota Bala Ayyappa,3Gowrabathuni Vidyasri,4Ganji L Venkata Naga Siva SAI Ram

Title : A Web-Based Approach for Seamless College Event Planning

Volume/Issue : 2026;03(06)

Page No : 1156-1161

This project studies user interaction patterns such as app usage, screen on time, and in app activities to find out what makes users satisfied. By analyzing large amounts of data, the framework identifies the most important factors that influence user experience and loyalty. The insights gained help developers create personalized experiences, improve app efficiency, and make data driven decisions to keep users engaged for longer. an intelligent system to classify mobile user behaviour into distinct categories like “Light user”, “Moderate user”, “Heavy users”, “Gamer” and “Social Media Addict”.

Keywords : Mobile App Analytics, User Behaviour Analysis, Machine Learning, Random Forest, Decision Tree, Support Vector Machine (SVM)

Author : 1Mrs. D. Madhuri,2Kagitha Monasri,3Amara Sri Sai Sudha Amrutha Varshini,4 Jaggarapu Venkat

Title : Comprehensive Analysis of Music Streaming Trends and User Engagement Patterns Using Spotify Data for Insights into Listener Behavior

Volume/Issue : 2026;03(06)

Page No : 1162-1167

OTP-Based Cardless Transaction System using Face Recognition for Secure Financial Operations allows users to perform ATM transactions without a physical card by using two-factor authentication, one-time-password (OTP) authentication with AI-based facial recognition to ensure multi factor verification of users during financial operations. Traditional card-based transaction systems are vulnerable to security threats such as card theft, skimming, duplication and pin leakage. Users may also forget or lose their cards. Therefore there is a need for a secure, reliable, and cardless transaction system that ensures strong user authentication while maintaining ease of use. In this project Python, Opencv, Mysql, Technologies are used and face detection , recognition models are used. It is provide a secure and convenient method for financial transactions without using physical cards.

Keywords : Python, OpenCV, Flask, MYSQL.OTP Cardless, Face Recognition

Author : 1Dr. A. Tirupatiah,2Busam Rani,3Akkala Lalitha Sri,4Mindala Poojitha

Title : OTP-BASED CARDLESS TRANSACTION SYSTEM USING FACE RECOGNITION FOR SECURE FINANCIAL OPERATIONS

Volume/Issue : 2026;03(06)

Page No : 1168-1173

Concerns about the validity of remote exams have grown as a result of the quick growth of online learning. The methods of manual invigilation are not scalable, costly, and prone to errors. The AI-driven real-time proctoring system described in this study uses computer vision techniques to autonomously monitor examinee conduct during online exams. The system examines real-time webcam video streams to identify anomalous head movements, multiple people, missing faces, mismatched identities, and the use of forbidden objects like cell phones. Real-time alerts and comprehensive activity logs are produced by combining YOLO-based object detection models, head pose estimates, and deep learningbased face identification. The suggested solution is appropriate for large-scale online assessments since experimental findings show that it successfully improves exam security while reducing human interaction.

Keywords : Online Proctoring, Computer Vision, Face Recognition, Head Pose Estimation, YOLO, Artificial Intelligence, Exam Security

Author : 1Dr. G. Prasuna,2 Shaik Nageena Jasmine,3Gampala Santhosh,4Karna Yoga Deepika

Title : AI-DRIVEN REAL-TIME PROCTORING SYSTEM FOR ONLINE EXAMS USING COMPUTER VISION

Volume/Issue : 2026;03(06)

Page No : 1174-1179

Artificial Intelligence and Computer Vision have significantly influenced modern beauty technology by enabling machines to analyze human facial features and provide personalized aesthetic enhancements. This project presents an Intelligent Makeup and Hairstyle Suggestion System that uses image analysis to detect face structure, skin tone, facial features, and hair attributes to recommend suitable makeup styles and hairstyles. The proposed system applies deep learning, face detection models, and imageprocessing techniques to generate accurate, user-specific beauty recommendations. This system aims to enhance user experience by reducing manual trial-and-error, improving accuracy, and providing real-time visualization.

Keywords : Computer Vision, Deep Learning, Face Detection, Makeup Recommendation, Hairstyle Analysis, Machine Learning.

Author : 1Dr. M. Ratna Raju,2Donda Yogasree,3Godavarthi Siva Kalyan,4Godavarthi Hema Krishna Priya

Title : INTELLIGENT MAKEUP AND HAIRSTYLE SUGGESTION THROUGH IMAGE ANALYSIS

Volume/Issue : 2026;03(06)

Page No : 1180-1186 |

Satellite image-based weather prediction using deep learning to make climate monitoring and forecasting more accurate. Traditional weather models rely on physical and numerical models, which are costly to run and not always reliable during extreme weather. The new datadriven approach learns atmospheric patterns directly from large amounts of satellite images. Convolutional Neural Networks help identify features like cloud patterns and weather structures. CNN and LSTM models are used to track changes in the atmosphere over time. The system is trained on both historical and current satellite data to make predictions more reliable. It helps forecast rainfall, cloud movement, and temperature changes. This deep learning method cuts down on computational costs compared to old models. It also works better when the weather changes quickly. Overall, the system helps with early warning, disaster response, and building long-term climate resilience.

Keywords : Satellite images, Weather prediction, Deep Learning, CNN, Climate Monitoring, Forecasting System.

Author : 1Dr. G. Prasuna,2 Jampani Sumasri,3Burla Venkata Srinivas Mahesh,4Galinki Mahima Tejaswini

Title : Satellite Image-Based Weather Prediction and Analysis Using Deep Learning for Climate Monitoring and Forecasting

Volume/Issue : 2026;03(06)

Page No : 1187-1193

To help healthcare managers, data holders, and data analzers by making hospital admission management more effective and data-driven, a web-based system called "Using Data Mining to Predict Hospital Admissions from the Emergency Department" was created. It ensures that only authorized and authenticated users can access system functionalities by offering secure login and registration options. Data Analyzers can access datasets to do analysis and produce admission predictions, while Data Holders can enter, update, and manage patient emergency information using an easy-to-use interface. Healthcare servers are capable of monitoring hospital data, reviewing prediction results, and assisting with operational decision-making. The solution incorporates a centralized database using MySQL to guarantee the safe and scalable storage of private medical data. The Data Analyzer dashboard allows for the visualization and study of prediction results, while the Data Holder dashboard makes patient data maintenance easier. Administrators have the option to create reports, keep an eye on system activities generally, and guarantee the prediction process's dependability and openness.

Keywords : safe login, user authentication, data mining, web-based application, and hospital admission prediction.

Author : 1Mr. N. Lakshmi Narayana,2Duddu Elisha,3Devarampati Venkata Pradeep,4Gandikota Janaki Ramaiah

Title : Using Data Mining to Predict HospitalAdmissionsfrom the Emergency Department

Volume/Issue : 2026;03(06)

Page No : 1194-1200

The Intelligent Forensic Analysis for Deepfake Video Detection using AI project tackles the rising problem of synthetic and altered videos that can be used for identity fraud, disinformation, and cybercrime. When it comes to advanced deepfake creation technologies, human inspection methods and traditional forensic approaches are frequently laborious, prone to errors, and ineffectual. In order to automatically evaluate video information and identify minute visual, temporal, and physiological irregularities that point to deepfake manipulation, this system makes use of artificial intelligence and deep learning techniques. By utilizing automated feature extraction, frame-level analysis, and trained neural network models, the proposed system ensures accurate, reliable, and scalable deepfake detection.

Keywords : Deepfake Detection, Intelligent Forensic Analysis, Artificial Intelligence, Machine Learning, Video Forensics, Deep Learning, Face Manipulation Detection, Multimedia Security, Digital Forensics, Synthetic Media Analysis.

Author : 1Dr. P. V. S. Sarma,2Alladi Harika,3Bondu John,4Kathi Daniel Christopher

Title : Intelligent Forensic Analysisfor Deepfake Video Detection Using AI

Volume/Issue : 2026;03(06)

Page No : 1201-1208

Fatigue is one of the leading causes of accidents in transportation and industrial environments, often resulting from prolonged working hours, monotonous tasks, and reduced alertness. Continuous monitoring of human alertness is essential to prevent fatigue-related accidents and ensure operational safety. This paper presents a real-time fatigue detection system based on facial feature analysis using artificial intelligence and computer vision techniques. The proposed system analyzes eye behavior using facial landmark detection and Eye Aspect Ratio (EAR) computation to identify signs of drowsiness such as prolonged eye closure. The system operates in real time using a standard webcam and provides immediate visual and audio alerts when fatigue is detected. Experimental results demonstrate that the system achieves high detection accuracy with low latency, making it suitable for deployment in transportation and industrial environments.

Keywords : Fatigue Detection, Facial Landmarks, Eye Aspect Ratio (EAR), Computer Vision, Artificial Intelligence, Real-Time Monitoring

Author : 1Mr. K. Ramesh Babu,2Devarakonda Lakshmi Tirupathamma,3Chelli Jhansi,4Gudisa Yaswanth

Title : Fatigue Detection from Facial Features Using AI to Prevent Accidents in Industrial and Transport Sectors

Volume/Issue : 2026;03(06)

Page No : 1209-1216

By allowing users to recognize meals and calculate calories, the Computer Vision Based Food Calorie Estimation system is a web-based platform that makes nutritional tracking easier. Conventional calorie tracking techniques take a lot of time and are not automated. Users can upload food photos for analysis with ease thanks to this clever system's useful solution. Predicted food labels and calorie data are visible to users. The model can quickly and reliably identify objects and calculate calories. TensorFlow is used for food categorization tasks and Python Flask for backend processing in the system's development. HTML, CSS, and Bootstrap are used to create an interface that is easy to use. User data is handled safely thanks to security measures. The platform increases awareness and lessens manual labor. All things considered, the method increases productivity and encourages healthy lifestyle choices worldwide.

Keywords : Food Recognition, Calorie Estimation System, Computer Vision, Deep Learning Model, Flask Framework, Web-Based Application

Author : 1Dr. D. Nagesh Babu,2 Jalla Simhadri,3Gattu Hema Gayathri,4Gaddam Praveen Kumar

Title : Computer Vision-Based Food Calorie Estimation Model Using Deep Learning for Health Monitoring and Nutritional Analysis

Volume/Issue : 2026;03(06)

Page No : 1217-1223

Work-life balance has become an important organizational strategy that enhances the psychological wellbeing, productivity and motivation of employes. Organizations that successfully encourage work-life balance are likely to have motivated employes which improves performance at an individual and organizational level. The present study is an attempt to investigate the work life balance and employe motivation of Nabil Bank Limited Kathmandu, Nepal. Primary data was obtained from 45 employes by means of structured, closed-ended questionnaires. The study employed descriptive statistics and correlation analysis to determine the relationship between variables. Data entry, tabulation and initial analysis were done using Microsoft Excel while SPSS software was used for data processing and analysis. The findings indicate that work-life balance is affected by many factors like flexible work environment, work load, supervisor help and family friendly regulations. Also, the results show that there is a positive relation between work-life balance and employe motivation. This means that those employes who have a good work-life balance are more motivated and dedicated to their careers. The study concludes that work-life balance initiatives are most effective when accompanied by a positive corporate culture and a supportive management style. Both factors enhance employe engagement and increase overall organizational effectiveness.

Keywords : Work-life balance, employee motivation, workload management, supervisor support, familyfriendly policy

Author : Prem Bhadur Dasuni

Title : Work-Life Balance and Employee Motivation in Nabil Bank Limited

Volume/Issue : 2026;03(06)

Page No : 1224-1233