Your Portfolio

Karim Elkobrossy

Projects

Machine Learning / Data Science Projects

1) Data Pruning for Cost-Efficient Training in Tribological Analysis on Journal Bearing Systems -Rolls Royce- (Upcoming publication)

Literature Gap Filling: New algorithms to address existing gaps in instance reduction in regression:
- Novel multi-tasking autoencoder.
- Novel adaptive DBSCAN algorithm.
- Modification of an existing Selesup algorithm.
Algorithmic Comparison: Comparison between the algorithms along with random sampling and the full dataset.
Pruning Percentage Analysis: Implications of varying pruning percentages on regression autoencoder, random sampling, and the full dataset.
Applying deep learning to analytically solve Reynold's equation using G ̈umbel’s assumptions.

Acadmic Paper Jupyter notebooks

2) Predictive modelling for the forecasting of water pumps functionality (Tanzania case study)

A predictive modeling approach to accurately forecast the operational status of water pumps across Tanzania. Leveraging a robust dataset, the methodology integrates machine learning algorithms, including deep learning and gradient boosting, with a focus on addressing the challenges of imbalanced data through preprocessing and feature engineering strategies.

Acadmic Paper Jupyter notebook

3) A Comparative Study of Feature Selection Techniques

This project investigates the impact of various feature selection methods on the performance of machine learning models in the context of big data.

Acadmic Paper Jupyter notebook

Programming / Web Development Projects

4) AI-powered UNO game

AI-powered UNO GAME with approximately 1000 lines of code (Python OOP)
3 bots and 1 user for playing the game
Fully functioning game with smooth animations

Game

5) Police Workforce Website

A complete police workforce website using (HTML ,CSS, JS, PHP, MYSQL). A technical guide and a user guide are included as well.

Website