Machine Learning in Practice Crash Course
What is This Course About?
There are tons of good machine learning courses in the web. So why bother taking this course? The reason is that this course focuses on different parts from other courses. Specifically, this course is about the fundamental knowledge of machine learning; the insights of different algorithms; best practices in applications. On the other hand, we will not pay much attention to the theoretical part of machine learning; instead, we provide some resources such as links to other courses or chapters of textbooks for interested readers.
Textbook
There are books for interested readers.
- [PC] Pattern Classification (2nd Edition), by Richard O. Duda, Peter E. Hart, and David G. Stork, Wiley-Interscience, 2000.
- [PRML] Pattern Recognition and Machine Learning, by Christopher M. Bishop, Springer 2006.
- [ESLII] The Elements of Statistical Learning: Data Mining, Inference, and Prediction (2nd ed.), by T. Hastie, R. Tibshirani & J. Friedman, Springer 2009.
- [MLAPP] Machine Learning: A Probabilistic Perspective, by Kevin Murphy, The MIT Press, 2012.
- [DM] Data mining: concepts and techniques, by Jiawei Han, Jian Pei, and Micheline Kamber. Elsevier, 2011.
- [DL]Deep learning, by Goodfellow Ian, Yoshua Bengio, and Aaron Courville. MIT press, 2016.
- Supplementary readings will be available online.
Schedule
Week | Time | Topic | Video | Reading | Lecturer |
---|---|---|---|---|---|
1 | 5/4 Mon. | Course Overview & Introduction | Lec1 | [PC] 1; [ESL] 1&2; [PRML] 1 | Conan Hu |
2 | 5/10 Sun. | ML Problem Reframing & Generalization & Metrics | Lec2 | Google Course | Conan Hu |
3 | 5/24 Sun. | Bayesian Decision Rule | Lec3 | Domingos’s excellent paper; [PC] 2 | Conan Hu |
4 | 5/31 Sun. | Naive Bayes & Linear Regression Methods | None | [PC] 3 | Conan Hu |
5 | 6/21 Sun. | Bias-Variance & Overfitting | Lec5 | [ESL] 4.5.1, 12; [PRML] 4.1.7, 7 | Conan Hu |
6 | 9/11 Sun. | Linear Classification Methods | Lec6 | [ESL] 4.5.1, 12; [PRML] 4.1.7, 7 | Conan Hu |
7 | 11/22 Sun. | kNN & Decision Tree | Lec7 | [PC] 5.1~5.5 | Conan Hu |
8 | 12/13 Sun. | Ensemble Methods: Bagging & Boosting | Lec8 | Conan Hu | |
9 | TBD | Ensemble Methods: GBDT & XGBoost | Conan Hu | ||
10 | TBD | Neural Networks & Deep Learning | CS231N | Conan Hu | |
11 | TBD | Clustering & Dimension Reduction & Visualization | Conan Hu | ||
12 | TBD | Frequent Pattern & ML in Practice | Conan Hu | ||
13 | TBD | ML in Practice: Tools & Data & Feature Engineering | Google Course ML Advice | Conan Hu |
Where to Go?
Here we list some recommended courses for deeply learning. Remember, if you want to gain a deep understanding, you should finish the assignments.
Coursera Machine Learning: An introductory given by Andrew NG. We suggest starting from here.
Pick one from the following two:
- Stanford CS229 Machine Learing: There are videos in Youtube and Bilibili (2018 version).
- CMU 10-701 Intro to ML:There are videos in Youtube (2016 version).
As the homework of the above three courses either based on MATLAB or do not cover much in coding part. Here we suggest the assignments of our course in Zhejiang University.
For the most popular deep learning, I recommend the following courses:
Stanford CS231n:Deep learning for CV. We suggest starting from here for deep learning.
Stanford CS224n:Deep learning for NLP.
Berkeley CS285:Deep Reinforcement Learning.
Last Word
Welcome to give us any feedback or suggestion. We are eager to hear any of them. Also, welcome to suggest topics that you would like us to cover!