The Deep Learning Revolution

book 12

这本书应该是机械学习 Machine Learning 的兰亭集序!

早期的 Nips Conference 群英荟萃,华山论剑!去的都是鸿儒!谈笑的都是先贤!!!

不管是现在流行的 kaggle, coursea, udacity, udemy, mooc ….. 这些创始人 
书本上的 Van Nuvemann, Shannon, Hardy , Number theory professor…
外星人 Elon mask.
最惊喜的是有babara , 我超喜欢她的learning how to learning.

我们的qualify 是take home exam, more of how you see / approach problems rather than have a definite solution for it.
感觉把这本书搞懂了就qualify 随便写,随便过!

Machine Learning 可以分三个大方向, Deep Learning , CNN neutral network 那一套, Forecasting (model prediction), NLP, Natural Language Process. 

这本书为了qualify 肯定要反复读的,书里面提到过的modeling, 我把我学过的总结一下。


zSVM support vector machine: able to classify, but won’t be able to show the exact support vector coefficients.

Gradient Boost

Gradient Descent(作者超喜欢!)


Continuous function/ Spike 

reinforcement learning: future rewards

who processes the most data wins. Of course the winner is China.

every night I pray, dear Lord, let the equation be linear, noise be Gaussian , and the variables be seprable.

小时候不觉得这个Log 有什么好? 现在知道她妙在能 separate variables. 
化繁为简, 乘除都变成加减!

*Note : 书里有好多地方提到了Go, Alpha Go, 但都是韩国的… 

想棋魂了,想光亮双子星,楚赢, 方旭….. all the crew

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Social media & sharing icons powered by UltimatelySocial
%d bloggers like this: