Bootstrap Regression

Two bootstrap techniques to get confidence intervals for regression coefficients are the case/pair bootstrap method and the residual bootstrap method. Case Bootstrap   The case or sometimes called pair is the response variables paired with the predictor variables . For B bootstrap samples Take a sample with replacement of size from  the corresponding bootstrap sample will be . […]

Leave One Out Cross Validation

Cross validation is one of many metrics for estimating out-of-sample error for predictive models. There are many flavors of cross validation hold-out, k-fold, leave one out (LOOCV), etc. I whipped up a neat little visualization script in R to help understand LOOCV. In Figure 1 the visual that shows how LOOCV works. The model is trained on the […]

Are you thirsty for predictions? – h2o

H2O is an open source machine learning / prediction platform from www.h2o.ai. Code is hosted on github, here. APIs are available for R, Scala, Python, Java. A REST API is available for web services or applications.   R H20 Tutorial Installing h2o package.

Creating a simple gradient boost model.

Node(s) store h2o data […]

rJava Package Not Installing on Mint Linux 17

While I was running R on my Linux Mint 17.2 OS I ran into a problem installing the xlsx package from the cran repository.

I produced the following error:

Didn’t tell me much.   Then I tried to install the rJava package

I got two errors

After searching I found the […]