This project utilizes the Arabic Sign Language (ArSL) dataset to train a deep learning model, creating a computer vision application that assists people to foster inclusive communication.

This project uses a dataset of PT exercises to train a deep learning model, creating a computer vision application that assists patients as they perform rehabilitation exercises.

Predicts glycemic variability using wearable device data and machine learning. The project provides a non-invasive way to monitor blood glucose levels, aiming to improve diabetes and pre-diabetes management. Feature engineering includes time-series analysis, lag features, and interaction features. The model achieves strong clinical metrics and is deployed for interactive use.

Multi-model pipeline for the Tamra-Thon Hackathon, using satellite imagery and machine learning to address challenges in the date palm industry. Includes a "Farm Performance Score" (K-Means clustering), 7-day NDVI health forecasting (LightGBM), and anomaly detection (Isolation Forest). The solution enables farm ranking, proactive alerts, sustainability insights, and supply chain forecasting. Features advanced time-series engineering, weather integration, and interactive dashboards.

A machine learning project that uses NASA's TESS satellite data to discover new exoplanets. Employs XGBoost classification with IsolationForest for automated transit signal detection in stellar light curves. The project mirrors real scientific discovery with supervised training on confirmed exoplanets and prediction on unknown stars. Features comprehensive data processing pipeline including normalization, detrending, and systematic noise removal.

A comprehensive Flask-based personal finance assistant powered by machine learning. Features a Linear Regression model for budget predictions, LSTM neural networks for Bitcoin price forecasting, and time-series models for NVIDIA stock predictions. Includes expense tracking, savings goals management, and predictive financial insights through an intuitive web interface.

A machine learning-based IoT security system that detects malicious network traffic using the IoT-23 dataset. Trained on various attack scenarios including Mirai, Torii, and other botnets. Implements multiple classification algorithms (Naive Bayes, SVM, Decision Trees, CNN) with the Decision Tree model achieving 93.7% accuracy. Deployed as a Flask web application for real-time threat detection.

A comprehensive healthcare web application that predicts CKD stages (1-5) using machine learning and provides personalized nutrition recommendations. Built with Flask and trained on a balanced dataset of 60,000 samples. Features multiple ML algorithms including Logistic Regression, SVM, Decision Tree, Random Forest, and KNN. Includes a rule-based nutrition system that generates diet plans based on predicted CKD stage and user demographics.

Analysis of Saudi Arabia's business establishments in Q1 2023, categorized by size and economic activity, sourced from Monsha'at's Open Data Platform.

Analysis of a US retail superstore's data, creating a dashboard to track key metrics and provide insights to management. The dashboard gives an overview of performance and identifies areas for improvement.

A consultant-led analysis of an online retail store's data to identify key revenue drivers and provide strategic recommendations to the CEO and CMO from marketing and operations perspectives.

A full-stack web application built with Flask to serve as a central hub for higher education institutions, streamlining academic and administrative processes for students and administrators.