## Recently Published

##### Investing $10,000 in the S&P 500

Had you invested $10,000 on 1929-07-01, your portfolio would have been worth about $3,000 after 8.6 years. In stark contrast, if you invested $10,000 on 2007-11-01, your portfolio would have been worth about $19,000 after 8.6 years.

##### Carl's Presentation

RStudio workshop for Rmarkdown

##### Beyond statistical significance

p-value thresholds for statistical significance should be abandoned

##### Convert an Annual Transition Matrix to a Monthly Transition Matrix using R

Finding the p-th or n-th root of a matrix.

##### Reducing Hospital Readmissions

Example of logistic regression

##### How to cheat, lie, and steal to get a better model

How to cheat, lie, and steal to get a better model

##### Measuring Overfitting

Measuring overfitting using a simulated data set

##### Three Deadly Sins of Predictive Modeling

Three Deadly Sins of Predictive Modeling

##### Time Series Cross-Validation

Time series cross-validation using the caret package. There is a bug in the document. Do not use the R.squared values from caret, use RMSE instead.

##### Comparing functional forms

Even if two models are algebraically the same, the predictions from the two models are not necessarily the same. Although the predictions between the two equivalent models converge as the sample size increases.
Do not use p-values to select or reject variables.

##### Using caret

Modeling the probability of getting approved for a credit card.

##### Boosting Linear Models

Boosting Linear Models and an assortment of other models on the German Credit Card data set.