COMP4211 (Spring2018)

This tutorial series is perpared by Kit Yeung

This is a repository to store the tutorials materials for COMP4211 - Mahcine Learning in HKUST.

To download any single file in this tutorial, you may follow this link.

Or you can clone this repository by:

git clone https://github.com/ckyeungac/COMP4211_Spring2018.git

Tutorial Schedule:

1. Tutorial 1 (2018-02-09)
   • Introduction of the tutorial
   • Azura Machine Learning Studio
   • Introduction to Python
2. Tutorial 2 (2018-02-23)
   • Introduction of machine learning work flow
   • Introduction to Numpy
   • Iris classification problem in scikit-learn
     • Iris dataset
     • Naive Bayes classifier
     • Train-test split
     • Accuracy
3. Tutorial 3 (2018-03-02)
   • Sentiment Analysis in scikit-learn(Toy example)
     • Artificial neural network (ANN)
     • Text Vectorization
     • K-fold cross-validation (Only basic knowledge included, the detail implementation is left for your assignment 1)
   • Assignment 1
     • 20 newsgroups dataset
4. Tutorial 4 (2018-03-09)
   • Perceptron (Implement from scratch)
   • ADALINE
   • Create a standalone workspace in Azure ML Studio.
5. Tutorial 5, 6 (2018-03-16, 2018-03-23)
   • Basic TensorFlow
     • MNIST dataset
     • Single layer neural netowrk
6. Tutorial 7 (2018-04-06)
   • Implement CNN using TensorFlow
7. Tutorial 8 (2018-04-13)
   • SVM in scikit-learn
8. Tutorial 9 (2018-04-20)
   • SVM in Azura Machine Learning Studio
     • Credit risk prediction
     • Different evaluation measures apart from accuracy
       • Confusion matrix
       • Precision, recall, F1-score
       • Area under the ROC curve (AUC)
9. Tutorial 10 (2018-04-27)
   • Nonlinear SVM in scikit-learn
     • Kernel
       • Radial basis function (RBF)
       • Polynominal