ju

Assignment Submission and Hints for JU BCSE Sessional courses

Customer churn, where customers leave a service provider for a competitor, poses significant challenges for telecom companies. This project develops a predictive model using a dataset of 100,000 records with 100 variables, aiming to identify likely churners and provide actionable insights to enhance retention strategies

This machine learning model was developed for "House Prices - Advanced Regression Techniques" competition in Kaggle by using several machine learning models such as Random Forest, XGBoost and LightGBM.

How to build, train, and evaluate a Generative Adversarial Network (GAN) for generating fashion images. The project covers all essential steps, from model building to training and evaluation.

Predict of wine quality

ML model predicting house prices using Linear Regression on Kaggle dataset. MAE: $28,108 | Python • Scikit-learn • Pandas

Finetuining Gemma3n with FineTome-100k dataset

Building an AI-powered analytics tool that turns messy business data into clear insights. Because small businesses deserve good data tools too.

A Multimodal AI assistant for automated plant pathology diagnosis, featuring an interactive UI built with Gradio. Developed during my MSc AI at BSBI, the system utilizes a fine-tuned EfficientNet-B0 architecture for high-accuracy image classification, integrated with natural language processing to provide real-time disease identification.

Python iteration exercises demonstrating for loops, range sequences, loop control (break), nested loops, and counters across healthcare, banking, retail, and entertainment workflows.

This repository contains a comprehensive machine learning workflow to predict housing prices using the classic Boston Housing dataset. The project demonstrates data preprocessing, feature scaling, model training, and hyperparameter tuning using regularization techniques.

Machine Learning

Config files for my GitHub profile.