Advanced Machine Learning for Industrial Applications (AMaLIA) –current

In AMaLIA, the primary focus is on developing advanced AI and Machine Learning techniques for timely and high-impact applications in industry. This initiative is supported by the National Science Foundation (NSF) through the Center for Big Learning (CBL). As a Postdoctoral Researcher, I serve as the Co-PI across multiple work packages, contributing to the advancement of intelligent solutions in the industrial landscape.

Co-Botics Intelligent Cooperating Robots and Humans (Phase 1 and Phase 2)

The project focused on the application of advanced machine learning and pattern recognition methodologies for facilitating shared intelligent cooperation between robotic units and humans. Developed a novel multimodal technique with smart decision strategies for Co-Botics. The project received funding (230K Euros) and support from Business Finland, as well as from NSF-funded Center, Visual and Decision Informatics (CVDI).

Mental Health and Productivity Boosting in the Workplace (Mad@Work)

This ITEA Smart health project focused on detecting and mitigating poor mental health conditions, such as work stress and burnout, which have not yet resulted in a diagnosed mental health disorder. In total 19 partners from 5 different countries were involve in this project. In the Mad@Work project, I developed stress detection algorithms for different work settings.

Virpa-D

The objective of the Virpa-D project was to explore how the operations-based environment will be realized in the future workplace and how the environment promotes well-being at work. More than 30 companies with their subcontractors and five research institutes take part in the four Virpa projects. Developed machine learning-based satisfaction index for working spaces.

Audio Source Recognition

The project focused on enhancing the audio noise source identification system using a one-class classifier. My primary contribution involved developing a real-time machine learning model for accurately classifying audio events as target or outlier. This project, funded by Sensornet in the Netherlands, served as a crucial component of my MS thesis.

Noise-specific speech denoising

In this project, we improved speech denoising for telecommunication systems by using the speaker and noise-specific algorithms. Turk Telecom supported the project at the vision and pattern analysis lab at Sabanci University. We proposed and developed a new method named “Recognize and separate,” where the algorithm adapts itself to the environment by identifying the noise in the audio signal and then removing the noisy part.

Sentiment Analysis of Review Text

In this project, I deduced diverse functions to classify tweets based on their polarity, assigning scores ranging from -1 (very negative) to +1 (very positive). I conducted experiments using various classifiers and algorithms to address this regression problem. This project was undertaken as part of a Machine Learning challenge during my coursework at Sabanci University, Turkey.

Facility monitoring and management

During my undergraduate studies at FAST NUCES, I developed the Facility Monitoring and Management (FMM) system, allowing remote control of electronic equipment via mobile phones. This project, integral to my BS studies, demonstrated my skills in software development. I presented the FMM system at Imperial College London in 2012 during the London International Youth Science Forum, showcasing both technical proficiency and effective communication on an international platform during BS studies.

Course Projects during BS Studies

• Non-Contact Heartbeat Sensor:

Investigated the development of a capacitive sensor to detect human heartbeat rates without direct skin contact.

• Smith Chart Interface:

Developed a graphical user interface to enhance understanding and interaction with the Smith chart.

• C++ KIOSK Development:

Designed and implemented kiosk software in C++ for a specialized automatic electronic vending machine.