Weighting/Parameter Optimisation: The weighting and parameter optimisation module,
as shown in Fig. 3.2, will be used adjust the training algorithms in order to produce a better learner
RapidMiner Optimise Weights (Evolutionary) operator. This operator calculates the weights of the features of
our football data sets by using a genetic algorithm (GA). The higher the weight of an attribute, the more
relevant it is considered. According to [14], a GA is a search heuristic that mimics the process of natural
evolution. This heuristic is routinely used to breed useful solutions to optimisation and search problems.
Genetic algorithms are an aspect of a larger class of evolutionary algorithms (EA), which produce solutions to
optimisation problems using such techniques as inheritance, mutation, and selection, as well as crossover [14].
Parameter optimisation, another data mining tool that finds the optimal values for a set of parameters using an
evolutionary approach, actually yields better results. This tool will be used to adjust the learning rate,
momentum and other hidden expert parameters until an improved model is built.