Personal Project
Loan Approval Prediction
10/2024
In my Loan Approval Prediction project, after EDA and data preprocessing, I optimized the k-NN classifier with cross-validation, achieving 89.4% accuracy. I also examined the Logistic Regression model with cross-validation, reaching an accuracy of 88.5%; ROC and AUC analyses showed that k-NN performed better. By effectively handling categorical data and fine-tuning with the CatBoostClassifier, I achieved 95% accuracy, obtaining the most successful model.
Smart Office System
09/2023 - 02/2024
Developed a smart system using a Raspberry Pi and RFID tracking to monitor employee attendance. Additionally, the system automatically controls the fan and lighting using sensors to provide an optimized work environment.
CS50 Bus
01/2024 - 03/2024
Developed a backend program for bus users inspired by Eshot, a local bus company in Izmir, as part of the Harvard CS50 Introduction to Programming with Python course. Features include creating and deleting accounts, reloading credit, and paying for rides.