跳到导航 跳到搜索

Machine Learning

Here are some notes and resources of Machine Learning. I'm taking this course COMP 4360 at University of Manitoba in the winter term of 2012. The professor is Jacky Baltes.


Course Website

It will be cleaned when the course is ended. I'll find some time to copy the contents to here.

Online Resources


Decision Trees

ID3 Algorithm

To calculate entropy and information gain of the data.

Entropy(S) = -p \sum_{i=1}^{n} -p * log_2(p)

In Sage:

  • Define: f(a,b)=-a/(a+b)*log(a/(a+b),base=2)-b/(a+b)*log(b/(a+b),base=2)
  • Representation: f(a, b) |--> -a*log(a/(a + b))/((a + b)*log(2)) - b*log(b/(a + b))/((a + b)*log(2))


Candidate Elimination

Candidate Elimination Algorithm

Artificial Neural Networks

Threshold function

Types of Activation Functions

  • Threshold
  • Linear
  • Sigmoid