ASPG Menu
search

American Scientific Publishing Group

Research Feed

Found 3772 matches for "All Articles"
Journal of Artificial Intelligence and Metaheuristics | 2025

Real-Time Violence Detection in Smart Cities Using Lightweight Spatiotemporal Deep Learning Models

The rapid development of smart-city infrastructure and the complexity of urban environments increase the need for automated systems that detect violence in surveillance footage. This paper presents a deep-learning architecture for detecting violence and weapon-related activity in CCTV footage using the Smart-City CCTV Violence Detection (SCVD) dataset. MobileNetV2 extracts spatial features from video frames through TimeDistributed layers, and a stacked Long Short-Term Memory (LSTM) network models temporal dependencies. The system processes 15-frame sequences at a resolution of 128 × 128 pixels. Batch normalization and dropout are used to improve generalization and limit overfitting, while a sigmoid output produces binary classifications. On the SCVD evaluation, the model achieved 99.58% accuracy and a binary cross-entropy loss of 0.0139. The Normal class achieved precision, recall, and F1-score values of 0.99; the Violence class metrics were reported as 1.00 after rounding. These results indicate the potential of compact spatiotemporal models for reliable surveillance and deployment on edge devices such as smart cameras and embedded systems.

groups
Muhammad Ahsan mail
link https://doi.org/10.54216/JAIM.090202

Volume & Issue

Vol. Volume 9 / Iss. Issue 2

Details open_in_new
Journal of Artificial Intelligence and Metaheuristics | 2025

Interpretable Rainfall Forecasting Using SHAP-Enhanced Machine Learning: A Case Study on U.S. Urban Climate Data (2024–2025)

Accurate rainfall prediction supports resilient communities, sustainable agriculture, water distribution, and disaster preparedness. Because rainfall depends nonlinearly on interacting meteorological conditions, conventional forecast-ing methods can be inadequate for short-term, localized events. This study develops a machine-learning framework for next-day rainfall classification using daily weather observations from 20 major U.S. cities during 2024–2025. The dataset includes temperature, relative humidity, wind speed, precipitation, cloud cover, and pressure, together with a binary Rain Tomorrow label. Six classifiers are evaluated: Random Forest, Support Vector Machine (RBF kernel), k-Nearest Neighbors, Logistic Regression, Naive Bayes, and Linear SVM. SHapley Additive exPlanations (SHAP) are used to interpret model predictions and feature contributions. Random Forest reports the strongest results, with precision, recall, and F1-score of 1.00; RBF SVM and KNN achieve F1-scores of 0.97 and 0.94, respectively. Logistic Regression, Linear SVM, and Naive Bayes achieve F1-scores between 0.76 and 0.77. SHAP analysis identifies precipitation and humidity as the most influential variables, consistent with meteorological expectations. These results suggest that combining an ensemble classifier with interpretable explanations can support transparent rainfall forecasting for agriculture, urban planning, and emergency response.

groups
Khaled Sh. Gaber mail -
Mahmoud Elshabrawy Mohamed mail
link https://doi.org/10.54216/JAIM.090203

Volume & Issue

Vol. Volume 9 / Iss. Issue 2

Details open_in_new
Journal of Artificial Intelligence and Metaheuristics | 2025

Multi-Classification of Brain Tumor MRI Images Hybrid VGG16 Support Vector Machine Model

Brain tumor research is essential for timely detection and appropriate treatment. Diagnosis is challenging because tumors vary in morphology, size, location, surface texture, and visual appearance across medical images. This study proposes a combined method for brain tumor detection from MRI images using three publicly available datasets containing 3,966 T1-weighted contrast-enhanced magnetic resonance images (T1-w MRI), classified as glioma, meningioma, pituitary tumor, or no tumor. The diagnostic pipeline applies preprocessing and data augmentation to improve image quality and variability. It then uses a VGG16 deep convolutional neural network to extract features and a Support Vector Machine (SVM) classifier to predict the class. The modified VGG16 output is supplied to the SVM, producing a test accuracy of 97.2%. Results from XGBoost, AdaBoost, Decision Tree, and K-Nearest Neighbors classifiers show that the SVM endpoint classifier performs best among the evaluated alternatives.

groups
Asifa Iqbal mail
link https://doi.org/10.54216/JAIM.090204

Volume & Issue

Vol. Volume 9 / Iss. Issue 2

Details open_in_new
Journal of Artificial Intelligence and Metaheuristics | 2025

LightGBM-Driven Earthquake Magnitude Prediction: A Comparative Machine Learning Framework Using Global Seismic Data

Earthquakes are among the most destructive natural hazards, causing severe damage to communities, loss of life, and major economic consequences. This study evaluates machine-learning models for magnitude prediction using an open-access global earthquake dataset from 2023. Five regressors are compared: Light Gradient Boosting Machine (LightGBM), Support Vector Regression (SVR), k-Nearest Neighbors (KNN), Ridge Regression, and Extra Trees. Training uses stratified cross-validation and hyperparameter optimization. The evaluation covers Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Bias Error (MBE), the coefficient of determination (R2), Nash–Sutcliffe Efficiency (NSE), Willmott Index (WI), Pearson’s correlation coefficient (r), and Relative RMSE (RRMSE). LightGBM performs best in the reported comparison, attaining MSE = 0.0474 and R2 = 0.9241. Its leaf-wise tree construction, scalability, and regularization support generalization to unseen data while maintaining computational efficiency. The results indicate that LightGBM can capture patterns in high-dimensional seismic data and provide a reproducible analytical framework for earthquake-prone regions. In combination with other advanced ML systems, such models may contribute to early-warning research, faster emergency response, and better-informed risk planning. The use of an open dataset and documented modeling steps also supports transparency and collaborative research.

groups
Nima Khodadadi mail
link https://doi.org/10.54216/JAIM.090205

Volume & Issue

Vol. Volume 9 / Iss. Issue 2

Details open_in_new
Fusion: Practice and Applications | 2025

A comprehensive and systematic exposition on Automatic Text Summarization Technique: A deeper coverage on extractive, abstractive and hybrid methods

Artificial Intelligence's remarkable advancement and Natural Language Processing enabled innovations that fulfill various vertical requirements. News summarization has become a popular topic where systems extract valuable semantic content and generate shorter abstracts from the original content. News readers benefit from a quick understanding of essential details because an informative summary provides them with important points without forced reading of the whole article. This article covers essential NLP news summarization methods, including Abstractive summarization, Extractive summarization, and Hybrid summarization, together with recent datasets, evaluation metrics, applications and future challenges. The main benefit of this work serves both researchers by providing them with complete information about contemporary summarization developments to select suitable summarization models during application development.

groups
Sini Raj Pulari mail -
Umadevi Maramreddy mail -
Shriram K. Vasudevan mail
link https://doi.org/10.54216/FPA.200114

Volume & Issue

Vol. Volume 20 / Iss. Issue 1

Details open_in_new
Prospects for Applied Mathematics and Data Analysis | 2024

HyperRough Cubic Set and SuperhyperRough Cubic Set

Rough sets provide a mathematical framework for approximating subsets using lower and upper bounds determined by equivalence relations, effectively modeling uncertainty in classification and data analysis. These foundational concepts have been further extended to structures such as Hyperrough Sets and Superhyperrough Sets. In this paper, we introduce the definitions of Hyperrough Cubic Sets and Superhyperrough Cubic Sets, and explore their fundamental properties. We hope that these developments will promote further research into applications such as decision-making based on Rough Set Theory and its extensions.

groups
Takaaki Fujita mail
link https://doi.org/10.54216/PAMDA.040103

Volume & Issue

Vol. Volume 4 / Iss. Issue 1

Details open_in_new
Prospects for Applied Mathematics and Data Analysis | 2024

Hyperalgorithms & Superhyperalgorithms: A Unified Framework for Higher-Order Computation

An algorithm is a finite, well-defined computational procedure that transforms inputs into outputs through a structured sequence of steps, guaranteeing termination and correctness. A multialgorithm comprises multiple algorithms augmented with a selection mechanism that dynamically chooses the most appropriate procedure based on input characteristics or contextual conditions. While these concepts have deep roots in computer science and beyond, this paper introduces two novel generalizations: the Hyperalgorithm and the Superhyper- algorithm. By leveraging the mathematical frameworks of hyperstructures and superhyperstructures, respectively, we extend the classical notion of computation to higher-order operations on sets and iterated powersets. We present formal definitions, illustrative examples, and a preliminary analysis of their computational properties, laying the groundwork for a unified theory of higher-order algorithms.

groups
Takaaki Fujita mail
link https://doi.org/10.54216/PAMDA.040104

Volume & Issue

Vol. Volume 4 / Iss. Issue 1

Details open_in_new
Prospects for Applied Mathematics and Data Analysis | 2024

Rethinking Strategic Perception: Foundations and Advancements in HyperGame Theory and SuperHyperGame Theory

Mathematical structures can generally be extended into Hyperstructures and SuperHyperstructures by leveraging powerset and n-th iterated powerset constructions (cf.7, 17, 31). These frameworks are particularly effective for representing hierarchical systems across various conceptual domains. Game Theory is a mathematical discipline for analyzing strategic interactions among rational agents with conflicting or cooperative objectives and finite choices.5, 10, 26 HyperGame Theory extends this by modeling situations in which players possess misperceptions or differing beliefs about the game being played.23 These ideas can be further generalized into the concept of SuperHyperGames.15 This paper explores the mathematical properties and illustrative examples of both HyperGame Theory and SuperHyperGame Theory. We hope that this investigation contributes to future developments in the theory and application of game-theoretic frameworks.

groups
Takaaki Fujita mail
link https://doi.org/10.54216/PAMDA.040201

Volume & Issue

Vol. Volume 4 / Iss. Issue 2

Details open_in_new
International Journal of Artificial Intelligence and Education Technology | 2024

Immersive Learning with the Metaverse’s Environment to Increase Academic Success and Motivation in learning Arabic as a Second Language for Non-Native Speakers

The metaverse's environment offers a unique opportunity for immersive learning experiences that can enhance education in ways never before possible. By creating virtual environments that simulate real-world scenarios, students can actively engage with the material and practice their skills in a safe and controlled setting. This technology has the potential to revolutionize the way we learn, making education more interactive, engaging, and effective for students of all ages. The integration of the metaverse's environment into Arabic language learning can provide non-native speakers with a more engaging and interactive learning experience. By creating virtual environments that simulate real-life situations, students can practice their language skills in a more realistic and practical way. The participants were 60 learners from non-native speakers enrolled in an Arabic Language course for intermediate level in the Arabic Language Center for Non-Native Speakers at the faculty of education at Mansoura University. The findings of research found that the immersive approach could help increase students' motivation to learn Arabic as a second language, leading to greater academic success in the subject. Additionally, the use of the metaverse can also help bridge the gap between language learners and native speakers, providing opportunities for real-time communication and cultural exchange.

groups
Reham Mohamed Al-Ghoul mail -
Ramy Samir Mohammed ALSeragy mail
link https://doi.org/10.54216/IJAIET.030101

Volume & Issue

Vol. Volume 3 / Iss. Issue 1

Details open_in_new