Passion is like genius; a miracle.

http://en.wikipedia.org/wiki/Transformation_matrix In linear algebra, linear transformations can be represented by matrices. Most common geometric transformations that keep the origin fixed are linear, including rotation, scaling, shearing, reflection, and orthogonal projection

Classical linear model minimizes where is residual and called LS(Least Squares) or Sum of Least Squares. But it is not robust against outlier. Thus we need roboust regression methods. Least Absolute Deviation Seek to minimize . This method has breakdown point zero meaning that even a small number of outliers can damage the goodness of…

Nice vid. on Kendall’s tau and Spearman’s rho. Part 1. Part 2. Here’s another explanation: http://www.unesco.org/webworld/idams/advguide/Chapt4_2.htm. In most cases, these values are very similar, and when discrepancies occur, it is probably safer to interpret the lower value. More importantly, Kendall’s Tau and Spearman’s Rho imply different interpretations. Spearman’s Rho is considered as the regular Pearson’s…

This is nice vid. from how2stats.com. Part 1. Part 2.

http://www.snl.salk.edu/~shlens/pca.pdf The easiest explanation on PCA among several documents I’ve read. It’s written by Jonathon Shlens.

http://en.m.wikipedia.org/wiki/Folded_normal_distribution If X is a random variable from normal distribution, then |x| follows folded normal distribution. Folding can happen anywhere. But if the folding is done where pdf is 0.5, it’s called half normal distribution.

Let’s use caret to find out the better # of hidden nodes. In the below, I needed many data so that default sampling method, i.e., k-fold CV, can have enough data in it. (i.e., if k=5 or 10, how can we run k-fold using just 4 data rows?) We may choose to instantiate trainControl, but…

http://caret.r-forge.r-project.org/Classification_and_Regression_Training.html The caret package (short for Classification And REgression Training) is a set of functions that attempt to streamline the process for creating predictive models. The package contains tools for: • data splitting • pre-processing • model tuning using resampling • variable importance estimation as well as other functionality. There are many different modeling functions…

Permutation test is a way of getting p value using randomization without assuming a certain distribution of data. The basic idea is simple. Suppose that we want to see if y = ax + b + error holds where x is 0 or 1. In other words, we’re interested if mean of y differs depending…