Machine Learning

The most difficult part for me was the code required to compute the partial derivatives.

Eventually using Octave I was able to vectorize the code and compress into a line.

Several years of Java programming experience didn’t prepare me for this type of code dealing with matrix transpositions and calculations.

Cost function I was asked to minimize is

$J(\theta_0,\theta_1) = (\partial/\partial \theta_0 )1/m \sum_{i=1}^{m} ( h_\theta(x^{(i)}) - y^{(i)}) x^{(i)}$

Plots showing the result of computing the gradient descent

Contour Plot

I also learnt single and multi-variate Linear Regression which will help me with my Capacity Planning calculations.