The project explores the use of generative AI algorithms to produce music similar to classical compositions. The project aims to develop algorithms that can create music that is both pleasant and approaches the characteristics of classical music. The study focuses on five different generative AI models: LSTM, VAE, GAN, RNN-VAE-GAN, and GPT, each with its unique architecture and approach to music generation.

The study investigates the use of generative models, particularly GANs, in educational recommendation systems to improve personalized adaptive learning. By addressing the shortcomings of traditional "one size fits all" educational approaches, the research aims to understand the current implementation methods, develop effective models, and assess their performance.

The Standard Model explains the basic particles and forces in the universe but has limitations. New Physics models are being explored to address these limitations. The COMET experiment aims to study charged lepton flavour violation, searching for processes not explained by the Standard Model. This study focuses on calibrating the momentum using simulations and data analysis utilizing Pionic Pion Capture process in the detector.

The project aimed to improve dry bean classification for farmers using the data from image processing techniques. The study focused on feature importance and hyperparameter tuning to enhance classification model performance. Various machine learning models were evaluated and the feature importance analysis revealed key features for classification. A web application was developed for easy access, contributing to the future of agriculture and smart farming practices.

The project focuses on predicting used car prices using machine learning. With rising demand for automobiles, accurate price predictions are essential. Objectives include developing machine learning models for a used car recommendation system and identifying key factors influencing prices. Five regression algorithms were used and evaluated with RMSE. Feature importance analysis was conducted to identify the most influential variables on price predictions.

The project uses machine learning to predict house prices based on factors like square footage, bedrooms, and property condition. The goal is to develop a reliable price prediction model for marketing properties. Various models were evaluated using R-Squared and RMSE metrics to find the best one. Error analysis was also conducted to assess model accuracy and reliability.