Predictive Financial Analytics: A Machine Learning Approach

• I evaluated a wide range of machine learning models, from conventional algorithms to advanced boosting methods like LightGBM, and achieved promising results.
• For feature engineering, I applied dimensionality reduction techniques, and I further optimized LightGBM using Optuna, a hyperparameter optimization framework.

E-Commerce Analytics: Descriptive & Funnel Analysis

• A descriptive analysis of an e-commerce website dataset
• Applied a funnel analysis to the dataset

Recognizing Handwritten Digits by using Artificial Neural Networks

• Matlab Implementation of a Feedforward Neural Network on the MNIST Dataset

Gibbs Sampling & EM Algorithm Implementation

• Mandatory course project for Simulation Methods for Statistical Inference
• Applied Gibbs sampling and Expectation-Maximization algorithm to the Iris dataset

Stock Control System

• The user interface developed on MS Access allows users to effortlessly enter, delete, and retrieve information from the database

Master Thesis

• Developed an analytical approach to highlight the detrimental effects of gerrymandering on electoral fairness.
• Utilized advanced technologies, including CPLEX Java API, R, and QGIS, to support modeling and analysis.
• Developed and solved a Mixed-Integer Mathematical Model optimally for relatively small cases.
• Designed and implemented a tabu search algorithm for efficient solutions to large-scale cases.

Political Districting Problem

• Demonstrating the detrimental effects of gerrymandering with mathematical programming
• The model is implemented using the CPLEX Java API

Hosted on GitHub Pages — Theme by orderedlist