Build Your Machine Learning on the Rock

A three moths schedule for Machine Learning Foundations.

1. Introduction

Nowadays Machine learning and Deep learning are widely used in all subjects, both industry and research organizations take huge effort to apply these methods.

I major in Information Management and Information System for my undergraudate study, and had some research experience related to Natural language Processing. I’m quite excited to study in the area of AI, especially in NLP/CL，since there are so many great contributors and huge progressed they’ve achieved.

More is Less and Less is More!

I believe that no matter how fast this field is growing, the most important thing we can do is to build a good foundation if we want to dive into it.

2. Planning

Thanks to an UP(NLP从入门到放弃) share his great learning experience bilibili, I decided to follow his guide and start my 3 moths ML schedule.

There 6 modules in this guide for about 3 months.

2.1 Prerequisites

• Week 1: Python
• Week 2: Python for data analysis
• Week3: Math(Linear Algebra, Calculus, Probability)

2.2 ML Theory

• Book : Statistical learning methods by Hang Li

  Must-learn chapter

  • Introduction
  • Perceptron
  • Naive Bayes
  • Decision Tree
  • Logistic Regression and MaxEnt
  • Boosted Trees
  • XGBoost

• Reference Blog: Jianping Liu’s Blog

• Principle

  • Can derive formulas by yourself.
  • Don‘t read every chapter in this book except for the must-learns'.
  • Don‘t reinvent wheels

2.3 ML Practice

• Alibaba Cloud TIANCHI

  • Industrial steam volume forecast
  • Repeat purchase prediction for Tmall users
  • O2O Coupon Forecast
  • Ali cloud security malicious program detection

2.4 DL Theory

• CV - cs231n

  • P1-P22
  • HW1,HW2

• NLP - cs224n

  • P1-P5 & P8,P9,P11
  • HW

• Pytorch Learning

  • Bilibili 刘二大人
  • A repo

2.5 DL Practice

• NLP

  • Ali Cloud Tianchi: News Text Categories

• CV

  • Ali Cloud Tianchi: Streetscape symbol recognition

2.6 Interview Questions

• Book: The Quest for Machine Learning

3. Note

Every time I have finished some in this schedule, I’ll post it to my blog.