Some Machine Learning Projects Chaokun has done before in courses or contests, including Natural Language Processing, Computer Vision, and Data Mining.

Natural Language Processing

Baidu MRC Contest

Baidu MRC Contest is an Machine Reading Comprehension Contest held by Baidu, aiming at solving three main porblems in NLP: Knowledge Extraction, Question and Answer, and Machine Dialogue. Chaokun and his group tried to solve the second one (QA) using BiDaf and Bert. The main work includes “Data Processiing”, “Model Optimization”, “Training and fine tune”, “Ensemble” .etc.

FastNLP Development

FastNLP is an open-source NLP tool set which developed and maintained by Prof. Qiu Xipeng ‘s group. Based on the Baidu MRC contest, Chaokun and his group added the Squad Metric in FastNLP and developed an MRC module in FastNLP

Computer Vision

This is the project required by Computer Vision Course. It is also an Contest held in 2019/5, aiming at remove the raindrop in single image. In this project, Chaokun re-produced an GAN model proposed in “Attentive Generative Adversarial Network (CVPR 2018))”. Besides, he also did some analysis of this DL Model.

Data Mining

This is one of the course projects required by Data Mining Course. It is also an Kaggle Contest held in 2014. In this project, Chaokun trained CNN and RNN models to extract the text feature, and got better result than baseline.

Link

Most Projects and be found in my Github repositories