Alperen Görmez

Graduate Research Assistant at University of Illinois Chicago
  • Claim this Profile
Contact Information
us****@****om
(386) 825-5501
Location
Greater Chicago Area
Languages
  • Turkish Native or bilingual proficiency
  • English Full professional proficiency
  • Japanese Elementary proficiency
  • Persian Elementary proficiency

Topline Score

Topline score feature will be out soon.

Bio

Generated by
Topline AI

You need to have a working account to view this content.
Join now
You need to have a working account to view this content.
Join now

Credentials

  • Eastern European Machine Learning Summer School
    EEML
    Jul, 2022
    - Oct, 2024
  • Certificate of Accomplishment, EEE 493 - EEE 494 Industrial Design Projects
    Bilkent University
    Jun, 2019
    - Oct, 2024
  • Deep Learning Specialization
    Coursera
    Apr, 2018
    - Oct, 2024
  • Machine Learning
    Coursera
    Apr, 2018
    - Oct, 2024
  • Certificate of Participation: Signal and Image Processing Days
    TED University / TED Universitesi
    Oct, 2017
    - Oct, 2024
  • Certificate of Participation and Competition: Google Hash Code 2017
    Google
    Feb, 2017
    - Oct, 2024
  • Certificate of Excellence: Matlab Programming for Engineers
    Coursovie
    Aug, 2016
    - Oct, 2024

Experience

  • University of Illinois Chicago
    • United States
    • Higher Education
    • 700 & Above Employee
    • Graduate Research Assistant
      • Aug 2019 - Present

      • Utilized the neural collapse phenomenon in early exit semantic segmentation models to reduce computational cost by 23% while preserving the accuracy.• Examined the combined effects of early exiting, pruning and sparsity using PyTorch.• Worked on early exit neural networks, adaptive inference, and model compression to decrease the computational cost of deep learning systems by 50% while preserving the performance.• Experimented on efficient distributed neural network training.• Supervised undergraduate students on early exit, knowledge distillation and conditional computation research.• Participated in the following communities: EEML, tinyML, SNN. Show less

    • Graduate Teaching Assistant
      • Aug 2019 - Present

      • Taught the ECE/CS 559 - Neural Networks using PyTorch (2021 Fall, 2022 Fall).• Instructed students MATLAB for the ECE 311 - Communication Engineering course (2020 Fall, 2021 Spring, 2022 Spring).• Helped students in the ECE 317 - Digital Signal Processing I course (2019 Fall, 2020 Spring).

  • Roku Inc.
    • United States
    • Technology, Information and Media
    • 700 & Above Employee
    • Machine Learning Intern
      • May 2021 - Aug 2021

      • Worked on reducing the inference time of a CTR prediction model in the Advertising Engineering team. • Used mlpy for cross-feature generation and feature transformation, Apache Spark for big data processing and TFX for pipelining. • Increased AUC by 0.03 and matched the inference time requirements. • Experimented with early exit networks and knowledge distillation techniques using TensorFlow. • Worked on reducing the inference time of a CTR prediction model in the Advertising Engineering team. • Used mlpy for cross-feature generation and feature transformation, Apache Spark for big data processing and TFX for pipelining. • Increased AUC by 0.03 and matched the inference time requirements. • Experimented with early exit networks and knowledge distillation techniques using TensorFlow.

  • Aselsan
    • Türkiye
    • Defense and Space Manufacturing
    • 700 & Above Employee
    • Candidate Engineer
      • Feb 2019 - Jun 2019

      • Built neural networks in TensorFlow and classified the sounds received by a passive sonar. • Worked on the visualization of the data collected by ultrasonic sensors using Python and Julia. Found a faulty sensor by analyzing the data. • Implemented sonar signal processing algorithms in MATLAB on a Linux system for the acoustics signal processing department. • Built neural networks in TensorFlow and classified the sounds received by a passive sonar. • Worked on the visualization of the data collected by ultrasonic sensors using Python and Julia. Found a faulty sensor by analyzing the data. • Implemented sonar signal processing algorithms in MATLAB on a Linux system for the acoustics signal processing department.

  • Nagoya University
    • Japan
    • Higher Education
    • 300 - 400 Employee
    • Research Student
      • May 2018 - Jul 2018

      • Conducted a research on pattern recognition and anomaly detection under the supervision of Prof. Kenji Mase. • Conducted a research on pattern recognition and anomaly detection under the supervision of Prof. Kenji Mase.

  • Argela Technologies
    • Türkiye
    • Telecommunications
    • 100 - 200 Employee
    • Intern
      • Jan 2018 - Feb 2018

      • Wrote Python programs for Linux machines to transfer files automatically between servers. • Contributed to the DSL-LTE Bonding Project by writing Python scripts to monitor the internet speeds of customers. • Led the team in automated testing processes using Robot Framework. • Wrote Python programs for Linux machines to transfer files automatically between servers. • Contributed to the DSL-LTE Bonding Project by writing Python scripts to monitor the internet speeds of customers. • Led the team in automated testing processes using Robot Framework.

  • Lumos Laser
    • Türkiye
    • Appliances, Electrical, and Electronics Manufacturing
    • 1 - 100 Employee
    • Summer Intern
      • Jun 2017 - Jul 2017

      • Optimized and simulated the data transfer between a computer and an FPGA of a fiber laser system using VHDL. • Optimized and simulated the data transfer between a computer and an FPGA of a fiber laser system using VHDL.

Education

  • University of Illinois Chicago
    Doctor of Philosophy - Ph.D., Electrical and Computer Engineering
    2019 - 2024
  • Bilkent University
    Bachelor of Science - B.S., Electrical and Electronics Engineering
    2015 - 2019
  • 名古屋大学
    School of Informatics
    2018 - 2018
  • Ankara Atatürk Anadolu Lisesi
    95.35
    2011 - 2015

Community

You need to have a working account to view this content. Click here to join now