Jung Soh

Projects

Natural language processing


Character-level language model
(Recurrent neural network)
Word analogy by document similarity
(Word vectors)
Sentiment classification
(Word vectors)

Neural machine translation
(Attention mechanism)
Named entity recognition
(Transformer network)
Question answering
(Transformer network)

Computer vision


Semantic image segmentation
(U-Net)
Car detection for autonomous driving
(YOLO algorithm)
Face verification and face recognition
(FaceNet)

Alcapa vs non-alcapa classification
(Transfer learning with MobileNetV2)
Sign language recogntion
(Residual neural network)
Face mood classification
(Convolutional neural network)

Sign language recognition
(Deep neural network)
Cat vs non-cat classification
(Deep neural network)
Sign language recognition
(Convolutional neural network)

Art generation, music generation and speech recognition


Merging content and style
(Neural style transfer)
Music generation
(LSTM)
Trigger word detection
(Gated recurrent units)

About me

I have worked in the field of biological data analysis for the last 15 years, processing vast amounts of genomic data for humans and other organisms and performing statistical analysis. In my most recent enterprise position, I designed and built a medical diagnosis product, where prediction of disease onset by a machine learning model was the key idea. I am seeking a data science or machine learning role in other industries as well as biotech, where I can leverage my data analysis experience and computing skills to tackle AI problems.

I have solid foundation in computing skills required to excel in data science and machine learning. I have experience building target prediction pipelines using open-source tools, including Python, SQL, R, Java and ML/DL frameworks such as TensorFlow. I am also eagerly expanding my machine learning and AI skill sets.

My earlier work before biological data analysis was on computer vision, including object detection, video processing and face/gesture recognition, with about 13 years of experience. I am updating these foundational skills in the context of deep learning to be able to solve significant computer vision problems with deep neural networks.

I have experience managing several multi-disciplinary teams and collaborated with subject matter experts with diverse expertise, including statisticians, biologists, IT administrators, and project managers.

Education

Certifications

Publications

I have published extensively with my work on biological data analysis and computer vision. Most of my publications can be accessed through my Google Scholar.

Patents

I am an inventor on several issued US patents. The fields of invention include molecular diagnosis of disease, biometric identification (signature and face), camera motion detection, and content-based image retrieval.

Teaching

While at Graz University of Technology and earlier at the University of Calgary, I have developed material for and taught courses in Python/R/Bash scripting, bioinformatics labs, and special topics in bioinformatics.

More information