flo1166/regression_javaclass

This was part of the lecture "object oriented programming of heuristics" at the KU Ingolstadt.

Regression_JavaClass

General:

This code solves a regression problem. Two methods are implemented:

• for linear regression (KQS)
• for quadratic functions, transformed as a linear problem It returns the m, t (or quadratic: a, lambda) parameter and the residuals (y - y_hat).

Structure:

• datapoints = build object where a datapoint consist of a week and product price

• DataPoint = a simple datapoint with week and price
• DatePointOverview = a abstract class to have datapoints relating to each other via DataPoint class and x and y array seperated (for calculating reasons)

• mathfunctions = interface, which contracts the functions to implement (for global usage)

• Regression = the interface to contract classes which want to interact

• regression = two methods implemented in Java

• LinearModel = implementation of the linear problem (KQS)
• QuadraticFunction = implementation of the quadratic problem, with log transformation and then processed like a linear problem