Profile

Metehan Cekic, PhD

Applied Scientist, AWS

metehancekic [at] gmail [dot] com

Linkedin | Github | Scholar

I am currently working at AWS AI as an Applied Scientist.

I completed my Ph.D. in the Electrical and Computer Engineering department at UC Santa Barbara. I was fortunate to be advised by Prof. Upamanyu Madhow. Through my research experience, I have developed an interest in Deep Learning and its applications. I received my B.S. degrees in electrical & electronics engineering and physics from Bogazici University, Istanbul, Turkey, in 2017.

Recently, I have been interested in developing robust machine learning models, specifically robust against adversarial attacks. An exciting line of work I've been part of also includes robust Radio Frequency fingerprinting. I also have some experience with reinforcement learning (specifically the A3C algorithm); you can check the cool AI bot we've developed for the Turkish card game called "Batak" ("Spades").

My hobbies include writing about my humble knowledge about the fields I have experience in; thus, I will regularly write blog posts here.

A complete list of my publications is online, along with some of my code.

Recent News

[Oct 2022] We've presented our paper titled "Neuro-Inspired Deep Neural Networks with Sparse, Strong Activations" at the IEEE International Conference in Image Processing (ICIP), 2022, Bordeaux. We are thankful to the community for their valuable comments and suggestions.

[Sep 2022] I’m happy to share that I’m starting a new position as Applied Scientist at Amazon Web Services (AWS)

[Aug 2022] I successfully defended my Ph.D! [Presentation]

[Jul 2022] We've presented our papers [1, 2] at ICML, 2022, Baltimore. We are thankful to the community for their valuable comments and suggestions.

[Jun 2022] Our paper titled "Neuro-Inspired Deep Neural Networks with Sparse, Strong Activations" is accepted to the IEEE International Conference in Image Processing (ICIP).

[Jun 2022] Our paper titled "Early Layers Are More Important For Adversarial Robustness" is accepted as a workshop paper for the Thirty-ninth International Conference on Machine Learning (ICML), New Frontiers in Adversarial Machine Learning.

[Jun 2022] Our paper titled "Layerwise Hebbian/anti-Hebbian (HaH) Learning In Deep Networks: A Neuro-inspired Approach To Robustness" is accepted as a workshop paper for the Thirty-ninth International Conference on Machine Learning (ICML), New Frontiers in Adversarial Machine Learning.

[Jan 2022] Our paper titled "Self-supervised Speaker Recognition Training Using Human-Machine Dialogues" is accepted to the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP).

[Nov 2021] We've presented our paper titled "Wireless Fingerprinting via Deep Learning: The Impact of Confounding Factors" at the 55th Asilomar Conference on Signals, Systems, and Computers, 2021, Asilomar. We are thankful to the community for their valuable comments and suggestions.

[May 2021] I am excited to announce that I will be joining the Alexa speech team at Amazon as an Applied Scientist Intern over Summer 2021.

[Dec 2020] I've recently talked to a Turkish Podcast about machine learning and artificial intelligence.

How do machines learn? What does the human brain have to do with artificial intelligence? If you are curious about these questions and more information, this episode is for you!


[Oct 2020] Successfully passed my PhD candidacy exam! [Presentation]

[Dec 2019] We've presented our paper titled "Robust Wireless Fingerprinting via Complex-Valued Neural Networks" at Globecom, 2019, Hawaii. We are thankful to the community for their valuable comments and suggestions.

[May 2018] I've been awarded the Outstanding Electrical and Computer Engineering Teaching Assistant (TA) award. I feel very fortunate to be a TA of smart and talented UCSB students.

Gallery

Here is my life in general: I really like walking in my free time, and like exploring the places around me. Generally, I like being socially active, getting to meet new people.