Algorithm Methodology

This document introduces a suggested approach to writing algorithms.

First of all, it is essential to note that there are different approaches based on the type of algorithm we are dealing with; Classic Machine Learning algorithms, Statistical methods and algorithms with cross-validation.

For statistical methods, there are not a lot of predefined steps. The general rules apply here; write the tests, produce the expected output, and write the algorithm. To write the algorithm, one has to analyze the method and create the steps specifically for each statistical method.

For Machine Learning algorithms, the same general rules apply. To write the algorithm there are two main function that should be implemented; fit and predict. (Jason)

For cross-validation algorithms, the implementation is based on the non-cross-validation algorithm.