AI and ML Applied to QA Automation
Latent Variable Modeling in Multi-Agent Reinforcement Learning via Expectation-Maximization for UAV-Based Wildlife Protection
Improved Blood Glucose Control using Machine Learning Algorithms
Deep Learning-Based Bharatanatyam Mudra Recognition using LSTM and GRU Architectures
Intellifusion Adaptive Decision Engine (IADE): A Hybrid AI Framework for Stock Market Forecasting
Artificial Intelligent Based Predictive Tool for Diabetes Prevention Management
A Smart Alumni Connect: Empowering University Networks with ML Driven Mentorship and Collaboration
Latent Variable Modeling in Multi-Agent Reinforcement Learning via Expectation-Maximization for UAV-Based Wildlife Protection
Advancing Autism Spectrum Disorder Diagnosis through Ensemble Learning
Enhancing Chronic Kidney Disease Prediction Performance through Algorithm Fusion: A Combined KNN - SVM Approach
A Comparative Exploration of Activation Functions for Image Classification in Convolutional Neural Networks
Machine Learning Solutions for the Healthcare Industry: A Review
The integration of Artificial Intelligence (AI) and Machine Learning (ML) in Quality Assurance (QA) automation is transforming software testing by enhancing accuracy, efficiency, and adaptability. AI-driven techniques, such as intelligent test case generation, self-healing automation scripts, and predictive analytics, enable faster defect detection and reduced maintenance efforts. ML models can analyze historical test data to optimize test execution, prioritize critical test scenarios, and identify patterns in software defects. Despite challenges such as data dependency, integration complexity, and the need for specialized expertise, AI and ML continue to drive innovation in QA automation. This paper explores the key applications, benefits, challenges, and future prospects of AI and ML in QA automation, highlighting their role in improving software reliability and development agility.
Protecting endangered wildlife from illegal poaching presents a critical challenge, particularly in vast and partially observable environments where real-time response is essential. This paper introduces a novel Expectation-Maximization (EM) based latent variable modeling approach in the context of Multi-Agent Reinforcement Learning (MARL) for Unmanned Aerial Vehicle (UAV) coordination in wildlife protection. By modeling hidden environmental factors and inter- agent dynamics through latent variables, our method enhances exploration and coordination under uncertainty. We implement and evaluate our EM-MARL framework using a custom simulation involving 10 UAVs tasked with patrolling protected habitats of the endangered Iranian leopard. Extensive experimental results demonstrate superior performance in detection accuracy, adaptability, and policy convergence when compared to standard algorithms such as Proximal Policy Optimization (PPO) and Deep Deterministic Policy Gradient (DDPG). Our findings underscore the potential of combining EM inference with MARL to improve decentralized decision- making in complex, high-stakes conservation scenarios. The full implementation, simulation environment, and training scripts are publicly available on GitHub.
The use of machine learning algorithms for harmless blood glucose monitoring is an exciting area of research that could revolutionize diabetes care. Traditional checking of blood glucose levels, like finger stick tests or continuous glucose monitors that require implantation, can be disagreeable and even painful for people. Continuous and painless monitoring is made possible by harmless technology that uses sensors applied to the skin. Machine learning techniques can analyze current information from these sensors to accurately predict levels of blood sugar. For accurate and reliable glucose readings, these algorithms must be optimized. This enables the development of models that can process data from several sources. This technology needs additional research and development despite its potential.
Deep learning techniques, particularly Recurrent Neural Networks (RNNs) with Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) architectures, offer effective solutions for recognizing Bharatanatyam mudras. These intricate gestures are essential for conveying the rich narratives and emotions inherent in this classical dance form. By training models on extensive datasets of labeled Bharatanatyam movements, the system can accurately detect subtle variations in gestures and expressions, significantly improving mudra recognition. This capability not only deepens practitioners' understanding of Bharatanatyam but also provides a valuable tool for storytelling during performances. Furthermore, it equips dance instructors with an effective method to demonstrate mudras to students, fostering a more engaging and interactive learning environment. Future extensions of this work could involve developing a system capable of real-time correction and verification of mudra positions, offering immediate feedback through display interfaces powered by deep neural networks. Overall, this research contributes to the preservation of Bharatanatyam as a digital heritage, ensuring its cultural significance endures in the modern era while enhancing both its educational and artistic value.
Forecasting stock prices is a complex and challenging task due to the volatile and unpredictable nature of financial markets. Traditional models frequently struggle with real-time data integration, precise sentiment analysis, and adaptability to dynamic market conditions. This paper introduces the IntelliFusion Adaptive Decision Engine (IADE), a comprehensive hybrid model integrating advanced technologies such as Deep Q-Learning (DQN), the Prophet Algorithm, Bidirectional Encoder Representations from Transformers (BERT), Adaptive Resonance Theory Neural Network (ART-NN), and transformer-based models with attention mechanisms. IADE aims to enhance user-friendliness, improve real-time forecasting accuracy, refine sentiment analysis precision, and provide adaptive predictive capabilities. The proposed system effectively improves forecasting accuracy and decision-making in volatile financial environments.
The global incidence of Diabetes mellitus is on the rise, leading to significant health challenges by increasing the pressure on healthcare systems. To enhance patient lower complications, early prevention is essential. This paper discusses the validation of an artificial intelligence (AI) prediction tool designed to identify individuals at a sensitive risk of diabetes. Advanced machine learning algorithms like random forest and support vector machine are used in this research, and neural networks are used to identify the patterns and relationships of diabetes onset, utilizing a comprehensive dataset that encompasses demographic, clinical, and elements from specific data sources such as electronic records and population surveys. Feature selection was implemented to improve the model's clarity and effectiveness.
Alumni are vital to a university's success, connecting the institution with the outside world. However, traditional alumni systems often fall short in fostering strong engagement, especially when it comes to mentoring current students. This paper introduces a Smart Alumni System (SAS) that combines social networking with machine learning (ML) and natural language processing (NLP) to enhance interactions between alumni, students, and faculty. This system supports a variety of activities such as mentoring, professional networking, group formation, and curriculum development. The ML algorithms provide personalized mentoring by matching people based on shared interests and skills, and NLP tools are used to analyze the student feedback towards the mentoring sessions. The existing system is a web application with networking features like connecting with friends, creating groups, and real-time messaging. By using ML and NLP in our system, it offers tailored recommendations, improves alumni relations, and makes mentoring more personalized and effective.
