Name: Chia-Ta Tsai
Type: User
Bio: Senior Applied Machine Learning in Wall Street, interested in #GenerativeAI, Deep Learning, #NLP, #RecommenderSystems, and Information Retrieval.
Twitter: ChiaTa_T29
Location: New York City
Chia-Ta Tsai's Projects
A logical, reasonably standardized, but flexible project structure for doing and sharing data science work.
Cookiecutter template for a Python package.
Docker images for dask
Batch dowload needed transofmrer models
A pipeline barebone for kaggle competition: Google Quest Q&A Labeling (Feb 2020)
11th place solution for Kaggle Human Protein Atlas Image Classification
7th-place solution code for https://www.kaggle.com/c/aptos2019-blindness-detection/overview.
This is a solution for the kaggle competition, ALASKA2, to detect secret data hidden within innocuous-seeming digital image.
This branch aims to refactor and modified U-Net training and perform stacking from repo of petrosgk/Kaggle-Carvana-Image-Masking-Challenge
Kaggling Home Credit Default Risk in a pipeline fashion.
Kaggle Kuzushiji Recognition: 2nd place solution
Kaggle Kuzushiji Recognition Solution
Ensemble of GLM and RF models for Kaggle Mercedes-Benz Greener Manufacturing Competition 2017
https://www.kaggle.com/c/recursion-cellular-image-classification
5th place solution for: RSNA Intracranial Hemorrhage Detection
ConvNet for Kaggle Statoil/C-CORE Iceberg Classifier Challenge, handling multi-inputs with augmentations on two channels
Deep Learning library for Python. Runs on TensorFlow, Theano, or CNTK.
A Keras implementation of CenterNet with pre-trained model (unofficial)
Keras/Ensemble learning for KKBOX Data Game - 17.06
A Library for Field-aware Factorization Machines
implementation of LinkNet in Keras
Modeling Popeline for SoftBank Forex Algorithm Challenge 2019
DeepThought's solution