Projects

Fall 2023

In this project, we aim to develop a system that can solve Sudoku puzzles using neural networks. The focus is on applying deep learning techniques, specifically convolutional neural networks (CNNs) with the Keras framework, to recognize patterns and deduce the correct numbers for each cell in the puzzle. 

Summer 2023

This project involves two key technological processes: the first converts input images into hand-drawn sketches, while the second uses Generative Adversarial Networks (GANs) to transform these sketches back into realistic images. 

This dual approach showcases advanced applications of deep learning, where the initial phase emphasizes feature extraction and pattern recognition to create sketch-like outputs, and the second phase leverages the adversarial training of GANs to fill in details, textures, and colors, turning simple sketches into lifelike images.

Fall 2022

Investigated Lyft’s riders’ data set by performing data wrangling, conducting exploratory data analysis, and building statistical machine-learned models, using python packages, to determine KPIs that guide riders’ cancellation decision 

Spring 2022

Combined logistic regression, with a decision tree, to build quantitative model, using python, for L'Oréal’s complex eye-tracking data, to predict psychological outcomes like fixation time, and recall effectiveness, with 89% accuracy

Fall 2020

Used multiple linear regression to create a predictive and statistical sales model optimized with log transformation. Experimented & validated large datasets with multiple psychological pricing techniques to recommend the ideal price.

Fall 2017

This project aims to enhance the level of immersion in computer games by creating an application that will allow users to interact with the game only using the gestures made by their hands. 

The proposed application enabled the user to play computer games without the need of either a keyboard or a mouse or any other expensive input devices. 

The users can play games through various gestures did by using only their hands. Therefore, the project focuses on replacing the traditional mouse and touchpads with human hand to interact with games which would allow more immersion of users and is also cheap to implement.