Machine Learning Los Angeles

Younwoo Roh

Los Angeles, California *********@*****.*** +1-213-***-**** www.linkedin.com/in/younwoo-roh/ Education

University of Southern California Los Angeles, CA

M.S. in Computer Science, GPA: 3.85/4.0 Aug 2023–Present Jeju National University Jeju, South Korea

B.S. in Computer Education, GPA: 3.91/4.0 (summa cum laude) Mar 2019–Aug 2023 Google Developers Community Jul 2022–Present

Google ML Bootcamp Jul 2022–Sep 2022

Experience

Mangoboost Seoul, South Korea

Machine Learning Systems Engineer intern Jun 2024–Aug 2024

• Accelerated inference speed of image ML models by 25% using Nvidia GDS (GPU Direct Storage)

• Adapted the FIO hardware benchmarking tool to support both AMD and Nvidia, as part of the DPU product for AMD

• Performed on Nvidia A40 and AMD MI100 GPU, to support company made DPUs (Data Processing Unit) for data centers USC Integrated Media Systems Center Los Angeles, CA Machine Learning Research Assistant Mar 2024–Present

• Working with Caltrans, implemented a 2-phase model to identify electric vehicles in IR camera videos using Yolov8 and Inception V3, achieving an F1 score of 0.96

• Optimized deep learning models, reducing inference time by 30% for real-time processing

• Submitted a paper on this project to the IEEE International Conference in Big Data (IEEE BigData) United States Air Force, Department of Defense Los Angeles, CA 96th Test Group Student Machine Learning Engineer Jan 2024–May 2024

• Automated high-frequency signal categorization on 100TB of random signals captured by the military

• Image classification approach using a transformer model after converting signals into normalized spectrograms

• Reduced 70% of the previous manual workload performed by electrical engineers within the Air Force Northrop Grumman Los Angeles, CA

Machine Learning Research Assistant Sep 2023–Dec 2023

• Automated fault detection for the company’s semiconductor chips, reducing manual workload by 60%

• Conducted detailed analysis of chips, identifying killer/benign defects, resulting in a 20% reduction in maintenance costs

• Identified defect existence with image classification, distinguished defects using ResNet18 for object detection

• Received the Best Presentation award at the 2023 USC DataFirst Fall event, sponsored by USC NanoCellect Biomedical San Diego, CA

Machine Learning Engineer Intern Apr 2022–Oct 2022

• Preprocessed laser-scanned 10µm scale cell images from flow cytometers to create a dataset for transfer learning

• Implemented a transfer learning model for classifying cell types and cycle stages for VERLO, an image-guided cell sorter

• Transferred and refactored existing Matlab code into Python, using TensorFlow and OpenCV Kakao Corp Jeju, South Korea

Software Engineer Intern Jan 2022–Feb 2022

• Developed and deployed an automated cryptocurrency transaction bot, achieving an average profit rate of 10% per week

• Implemented various stock trading strategies and provided backtesting results to users

• Automated recommendation and execution of optimal strategies using UpBit and Telegram API LG Soft India Bangalore, India

Machine Learning Engineer Intern Jul 2021–Aug 2021

• Implemented rotten food detection embedded in LG’s smart refrigerator based on webOS, achieving 90% accuracy

• Linked internal food storage to a database to automatically purchase replacements for rotten ingredients

• Awarded the global-scale Gold Award at the LG Soft India webOS OSE Hackathon, sponsored by LG Soft Publication

• (Under revision) Krish Sukhani, Jooyoung Yoo, Younwoo Roh, Tejas Rajurkar, Seon Ho Kim. (2024). Image Machine Learning Approach to Detect Electric Vehicles Using Infrared Camera. IEEE International Conference in Big Data (IEEE BigData).

Skills

Programming: Python, C, C++, C#, Java, JavaScript, HTML5, CSS3, SQL, PHP, JQuery, Ajax Frameworks & Tools: Tensorflow, Pytorch, YOLOv8, Scikit-learn, Keras, Pandas, Numpy, Matplotlib, Seaborn, Django, Transformer, ResNet, Enact, React

Contact this candidate