Problem 2
Part A
Below shows datasetA fitted with 3 Gaussians.
Below shows datasetB fitted with 3 Gaussians.
Part B
For part B of problem 2, for each possible K = 1 to 5, I calculated the and
Problem 4
For this question, first I tried different combination of polynomial order and C (the
upper bound in non-separable cases). The polynomial order varies from 1 to 5, and C
value varies from 1 to 1000000
Machine Learning Homework 1
Yuechen Zhao (yz2877)
September 30, 2015
Problem 1
In problem 1, I randomly split the dataset into two halves. Then I used the
provided code in polyreg.m to calculate the . Also I calculated the training error
and the testing e
COMS W4117
Compilers and Translators:
Software Verification Tools
Lecture 7: Data-Flow Analysis Frameworks
September 25, 2007
Lecture Outline
1.
2.
3.
4.
5.
6.
Review
Data-flow analysis frameworks
Iterative algorithm for general frameworks
Meaning of a da