Recently is a period of idle, do not want to waste, remember before there is a collection of machine learning link Andrew ng NetEase public class, of which the overfiting part of the group will report involved, these days have time to decide to learn this course, at least a superficial understanding.
Originally wanted to go online to check machine learning books, found that Lee's "statistical learning Method" and PRML (Pattern recognition and machine learning) is highly respected, have time to see it.
Then in the library met God bless, I recommended the Coursera this site, which has Andrew Ng for the network version of the machine learning tutorial, very good. The following notes are based on this course.
Https://www.coursera.org/course/ml
Week One:
A:machine Learning
Supervised learning:regression classification
Unsupervised Learning:cluster
and reinforcement learning, recommender systems
B:linear regression with one variable
Linear regression:
Hypothesis,cost function (why the denominator in the least squares estimate has a factor of 2), Contour plots (the values on one line in the contour chart are equal)
Gradient Descent:
Alpha:learning Rate
Ifαis too large, gradient descent can overshoot the minimum. It may fail to converge, or even diverge.
Gradient descent can converge to a local minimum, even with the learning rateαfixed.
Gradient Descent for linear regression:
Convex Function for it.
"Batch" Gradient descent:
Batch:each step of gradient descent uses all the training examples.
C:linear Algebra Review
If A is an M x m matrix, and if it have an inverse
(How to tell if a matrix exists without an inverse matrix)
Matrices that don ' t has an inverse is "singular" or "degenerate"
Machine learning notes-from Andrew Ng's instructional video