Search
106 articles for “feature engineering”
-
Prediction of Mechanical Properties for Advanced Engineering Applications utilizing Polymer Composite Materials by Machine Learning
Abstract: Polymer composites show great promise as engineering materials because of their mechanical performance, resistance to corrosion, lightweight nature, and adaptability in design. Aerospace, automotive, biomedical, maritime, and civil engineers all rely on mechanical property prediction to cut down on trial expenses, expedite product development, and optimize material selection. Speedy design optimization is not possible using traditional numerical and experimental methods due to the high costs associated with material characterisation, computational …
Published in Journal of Polymer & Composites · Vol. 14, Issue 3, 2026 Read article
-
A Comprehensive Analysis of Classification Methods for Churn Prediction in Financial Services
Abstract: Persistent issues that affect long-term revenue in the banking sector include excessive client attrition. Customary churn models depend on measures related to customer satisfaction, which often result in low predictive accuracy due to their subjective nature. This study proposes an effective early warning model to address customer churn in financial services. Data is preprocessed through cleaning, one-hot encoding, Z-score normalization, and Min-max scaling. To handle class imbalance, the SMOTE algorithm …
Published in Journal of Advanced Database Management & Systems · Vol. 12, Issue 2, 2025 · pp. 47–61 Read article
-
SpecForesight: A Predictive Analytics Pipeline for Laptop Price Forecasting
Abstract: This paper frames laptop pricing as a supervised predictive analytics problem, transforming product specifications into feature-rich signals to forecast price with calibrated regression models and operational guardrails against drift. A structured pipeline ingests tabular listings, performs data cleaning, and engineers domain-informed features (e.g., central processing unit (CPU) family and clocks, graphics processing unit (GPU) tiering, memory/storage density, display, and touch capabilities), followed by encoding and normalization to optimize model learnability. …
Published in Journal of Computer Technology & Applications · Vol. 17, Issue 1, 2026 · pp. 61–71 Read article
-
Machine Learning-Based Quantification of Polymer Structure Property Relationships for Predictive Material Design
Abstract: Polymer structures exhibit complex, hierarchical arrangements that strongly influence macroscopic properties, yet consistent quantification remains challenging due to nonlinear interactions and limited unified modeling strategies. Existing approaches inadequately capture generalized structure–property mappings across diverse polymer systems. This research aims to establish a machine learning-based quantification model for polymer structure–property relationships to support predictive material design. A Polymer Structure Property Dataset of 5,000 polymer samples includes structural descriptors and experimentally measured …
Published in Journal of Polymer & Composites · Vol. 14, Issue 3, 2026 · pp. 737–754 Read article
-
AI for Cybersecurity: Deploying Machine Learning for Network Traffic Anomaly Detection
Abstract: The growing sophistication of cyberattacks and the growth of network traffic necessitate sophisticated anomaly detection methods. This study overviews the use of artificial intelligence (AI) and machine learning (ML) to counter these challenges, as noted in current studies. It analyses supervised learning (SVM, Decision Trees), unsupervised learning (K-means, DBSCAN), and deep learning (CNNs, RNNs, Auto-encoders) approaches, considering their strengths and weaknesses. The research integrates current developments in AI/ML-based network anomaly …
Published in International Journal of Computer Science Languages · Vol. 3, Issue 2, 2025 · pp. 1–10 Read article
-
Timestamp Extraction and Log Classification Using Supervised Machine Learning: A Comparative Study
Abstract: In modern software systems, logs are vital for monitoring application behavior, diagnosing issues, and analyzing performance. Timestamps are especially important for sequencing events, identifying anomalies, and understanding system failures. However, detecting timestamps in logs is challenging due to inconsistent formatting across systems and the presence of timestamp-like strings in non-timestamp fields. Traditional rule-based methods often fail in such cases. This study proposes a supervised machine learning approach to accurately classify …
Published in Journal of Software Engineering Tools & Technology Trends · Vol. 12, Issue 3, 2025 · pp. 26–38 Read article
-
Data-Driven Energy Forecasting for Smart Homes: Ensemble Learning from IoT Meters and Relevance for Polymer-Composite Based Smart Infrastructure
Abstract: Reliable estimation of household electricity demand is relevant in creating efficiency in energy usage, optimization of the loads, and intelligent demand-side management in intelligent grid systems. This paper introduces a varied machine learning model that approaches residential electric consumption prediction using an assortment of ensemble regression boosts, including Linear Regression, Lasso Regression, Decision Tree Regressor, Random Forest, and Gradient Boosting, to predict residential electricity consumption environments on a time-series arrested …
Published in Journal of Polymer & Composites · Vol. 14, Issue 1, 2026 · pp. 29–64 Read article
-
Identifying and Implementing a Machine Learning Model Suitable for Processing Visually Evoked Potential
Abstract: A Brain-Computer Interface (BCI) is a system that translates brain activity patterns into computer commands, bypassing physical movement. Electroencephalography (EEG) is commonly used to acquire signals in BCI research. Visual evoked potentials (VEPs) are brain responses in the visual cortex to visual stimuli. Recent studies show that exposing individuals to flickering at a consistent frequency generates EEG signals synchronized with the stimulation. Efficient extraction of VEP signals begins with preprocessing …
Published in Journal of Microwave Engineering and Technologies · Vol. 11, Issue 2, 2024 · pp. 1–8 Read article
-
Machine Learning Approaches Towards Resume Classification
Abstract: Finding the right person for an open position can be an unnerving task, especially when there are many applicants, and if the recruiter or the Human Resources department must sort and further categorize all those resumes then it will be a labor-intensive, time-consuming, and tiresome task. Additionally, human assessment of resumes may be biased and prone to mistakes. Manually screening the proper candidate's resume from the pool is not practicable; …
Published in International Journal of Electronics Automation · Vol. 1, Issue 2, 2023 · pp. 1–7 Read article
-
Mechanical Strength Prediction of Nano-Silica Concrete Composites Using Machine Learning Techniques
Abstract: Nano-silica, or nanosilica, refers to silicon dioxide nanoparticles, which are a kind of silica (SiO₂) with diameters that often fall below 100 nanometers. This nanomaterial has attracted considerable attention because of its distinctive characteristics and diverse array of uses, notably in augmenting the performance of materials such as concrete. The integration of nanoparticles with cementitious matrix in nano-silica concrete offers a viable approach to improving the mechanical characteristics and longevity …
Published in Journal of Polymer & Composites · Vol. 13, Issue 1, 2025 · pp. 963–973 Read article
-
Pothole Detection utilising Machine Learning: A Review
Abstract: Potholes must be found and fixed quickly in order to maintain infrastructure, maximize transportation systems, and guarantee road safety. Using the Sequential API and the Keras library, this study presents a neural network model for pothole detection. Convolutional layers with ReLU activation, global average pooling, dense layers with dropout, and softmax activation for binary classification make up the model architecture. Image loading, resizing, array conversion, labeling, shuffling, normalization, and one-hot …
Published in Journal of Control & Instrumentation · Vol. 16, Issue 1, 2025 · pp. 35–43 Read article
-
Enhance Thermal and Conductive Properties through Graph Neural Network-Based Machine Learning-Driven Advanced Polymer Material Design
Abstract: Advanced polymer materials are widely used in modern engineering and manufacturing because of their lightweight nature, flexibility, durability, and adaptability to different applications. However, designing polymer materials with enhanced thermal and electrical properties remains a challenging task. The performance of polymers is influenced by a complex combination of molecular structures, filler materials, processing parameters, and nanoscale interactions. Conventional optimization methods often require extensive experimental trials and computational resources, making it …
Published in Journal of Polymer & Composites · Vol. 14, Issue 3, 2026 Read article
-
Monitoring of Unauthorized Identity and Access Behaviour for Outsourced Data in Cloud Environment
Abstract: The outsourcing of data is a significant challenge in the modern cloud computing ecosystem when it comes to tracking unauthorized identification and access behaviour. In order to overcome this issue, this research suggests a thorough method for reliable anomaly detection in cloud systems. Improving data security and offering a trustworthy monitoring system are the two main goals. The suggested approach proceeds methodically, gathering information from several sources such as user …
Published in Journal of Communication Engineering & Systems · Vol. 14, Issue 2, 2024 · pp. 9–19 Read article
-
Machine Learning Innovations for Effective Spam Comment Filtering in Social Networks
Abstract: The increasing prevalence of social media platforms has revolutionized communication, fostering unparalleled levels of connectivity and data exchange. However, the widespread increase in spam comments presents a serious threat to the integrity of online discussions, potentially undermining the quality of interactions. To confront this issue, our proposed model utilizes machine learning techniques to bolster spam comment detection across various social media platforms. This endeavor involves a thorough investigation encompassing data …
Published in International Journal of Algorithms Design and Analysis Review · Vol. 2, Issue 2, 2024 · pp. 19–24 Read article
-
A Comparative Study of Deep Learning Methods for Depression Detection in Social Media Data
Abstract: With the rise of social media platforms like Twitter, Reddit, and Facebook, individuals increasingly share personal information about their moods, behaviors, and mental states. This trend provides a unique opportunity to leverage large-scale textual data for understanding and monitoring mental health conditions, particularly depression, a prevalent and challenging mental health issue. Traditional depression assessments are often confined to clinical environments and lack the capacity for real-time monitoring. In contrast, social …
Published in Journal of Artificial Intelligence Research & Advances · Vol. 12, Issue 3, 2025 · pp. 55–65 Read article
-
Harnessing Machine Learning for Stock Movement Prediction: A Review of Current Approaches
Abstract: Stock price prediction is a crucial task in financial analysis, aiding investors and traders in making informed decisions. This study investigates the use of deep learning methods, particularly Long Short-Term Memory (LSTM) networks, for predicting stock prices based on historical market data. The dataset, sourced from Yahoo Finance, consists of time-series stock price data, which is preprocessed, feature-engineered, and visualized to improve prediction accuracy. The model's performance is assessed using …
Published in Journal of Advances in Shell Programming · Vol. 12, Issue 2, 2025 · pp. 29–40 Read article
-
A Machine Learning Approach to Forecasting Outcomes in Limited Overs Cricket
Abstract: This study explores the application of machine learning techniques to forecasting outcomes in limited overs cricket matches, with a particular focus on One Day Internationals (ODIs). The research investigates how classification algorithms can be effectively utilized to analyze both contextual and dynamic factors that influence match results, including venue details, toss decisions, team strength, and historical performance records. By employing a structured methodology encompassing feature selection, data preprocessing, model training, …
Published in Recent Trends in Sports · Vol. 2, Issue 2, 2025 · pp. 09–19 Read article
-
Enhancing Maintenance Decision-Making in Thermal Power Plants Using Generative AI-Based Fault Diagnosis
Abstract: The growing complexity of operation and power consumption of thermal power stations involve the need to have intelligent fault diagnosis systems that can be used to guarantee reliability and safety in operation. In this research, a Generative AI (GenAI)-based hybrid architecture of early fault detection and predictive maintenance is proposed to improve the decision-making process of the maintenance team. The data-driven analytic approach combines methods of data-driven analytics, Generative AI …
Published in International Journal of Energy and Thermal Applications · Vol. 4, Issue 1, 2026 · pp. 25–33 Read article
-
Aerodynamic Optimization of UAV Wings Using Machine Learning
Abstract: Unmanned Aerial Vehicles (UAVs) are increasingly deployed across defense, transportation, agriculture, and environmental monitoring, demanding improved aerodynamic efficiency to enhance endurance, stability, and payload capacity. Traditional aerodynamic optimization approaches, relying on computational fluid dynamics (CFD) simulations and wind tunnel experiments, are often time-consuming and computationally expensive. This study proposes a machine learning (ML)-driven framework for the aerodynamic optimization of UAV wing geometries, aiming to significantly reduce design cycles while improving …
Published in International Journal on Drones · Vol. 2, Issue 1, 2026 · pp. 1–7 Read article
-
CNN-Based Diagnosis of Skin Cancer from Dermoscopic Images
Abstract: Skin cancer has become one of the diseases widely spread over the globe, with melanoma becoming a severe threat to one’s health. Detection of such diseases at the initial stage saves an individual from drastic damage. Using a Convolutional Neural Network (CNN) for detecting skin cancer through image classification as benign or malignant provides significant support to dermatological practice and reduces dependence solely on subjective visual examination. Dermatologists often face …
Published in Research and Reviews: Journal of Oncology and Hematology · Vol. 15, Issue 1, 2026 · pp. 37–42 Read article