机器学习任务的state-of-art之github及个人总结

前言

深度神经网络其实更加适合做感知，而贝叶斯理论的核心是推理，只有从感知到推理才能到决策。所以最终来讲，你希望达到一种理性的推理、理性的决策，这里面正好是贝叶斯网络一个大行其道的地方—余凯

1. State-of-the-art result for all Machine Learning Problems

https://github.com/RedditSota/state-of-the-art-result-for-machine-learning-problems

2. Google Research

https://research.googleblog.com/

3. paperWeekly

http://www.paperweekly.site/collections/10/papers

4. paperswithcode

https://paperswithcode.com/sota

5.OCR

https://github.com/hwalsuklee/awesome-deep-text-detection-recognition

6. NLP

https://github.com/sebastianruder/NLP-progress