Drug Design with Artificial Intelligence Techniques – October 2021 – Present

This project aims to generate molecules with desired properties using generative models and optimization techniques. This project contributes to Moonshot project of the Japanese Government.

inzva – Google Machine Learning Bootcamp – August 2022 – December 2022

I am an organizer and the lecturer at the Machine Learning bootcamp which we train 130 participants in Turkey. Google organizes similar bootcamp across the world and our program gave the best metrics in terms of the student performance.

Image Classification on Mobile App – December 2021 – March 2022

For my client, I have developed an image classification algorithm for the mobile app available on both AppStore and GooglePlay. The project includes segmentation of the target object from the user input image and classifying the objects using CNN-based models and post-processing techniques.

Medical Image Segmentation – March 2021 – December 2022

Medical Image Segmentation problems usually suffer from a lack of annotated data. In this project, we use data augmentation techniques to improve the segmentation quality of the U-Net model. In addition, we are working on a Python library that selects augmentation techniques to be used and optimizes their parameters using Bayesian Optimization.

Customer Life Time Value Prediction & Campaign Optimization – January 2021 – August 2021

Three-year LTV of the customers are predicted using survival models for each cohort, and the appropriate testing mechanisms are implemented. Model results are used for campaign optimization and marketing decisions. Model is deployed into production using Airflow.

Meditation Recommender System – March 2021 – August 2021

The first version of the Meditation Recommender System at Meditopia. Collaborative filtering algorithms with implicit feedback are used, and the model is integrated using Airflow.

VarRedOpt: A Framework for Variance Reduction – May 2020 – December 2020

An R Library published on CRAN for efficient Monte Carlo Simulation techniques with built-in simulation functions for Option Pricing.

Recompy: A Framework for Recommendation – March 2020 – June 2020

In this project, we have created a Python Library, Recompy. It is a framework for matrix factorization

techniques and other baseline methods to be used to develop recommender system algorithms.

Anomaly Detection with Generative Models – August 2019 – November 2019

Fraud detection and cyber-attack detection using generative models like Variational Autoencoders

and GANs. The project report is available online.

Sketch-Based Face Analysis / Recognition – February 2019 – June 2019

Creating a face recognition algorithm that recognizes the identity of people using their sketch

drawings. HOG and CNN methods are used. AR and CUHK databases are used.

Algorithmic Trading Application with Trend Following Strategies – February 2019 – June 2019

In this project, we used Moving Average, HP Filter, and Lowess methods to create indicators to be

used for Trend Following Strategy.

Recommender System for TeknoSA — December 2018 – May 2019

In this project, I created and deployed a product recommender system for TeknoSA retail stores, one of Turkey’s most prominent electronic retail companies.

Detection of Denial-of-Service Attacks with SNMP/RMON — October 2017 – April 2018

In this study, a Denial of Service (DoS) attack scenario is examined using an Ethernet switch that

supports Remote Monitoring (RMON) and Simple Network Management Protocol (SNMP).