Decision trees as a machine learning with using genetic algorithms to evolutionary selection and construction for quality decisions

14/05/2015 - 12:00
Speaker: 
מיקום: 
Abstract: 

The Evolutionary-Random-Forest algorithm, basically chooses the best trees from the population, and then creates new, more accurate “child” trees from those “parents”. In effect, this creates a new population of trees with enhanced classification accuracy. The predictions are carried out using a small population of enhanced-accuracy “child” trees, thus increasing the prediction algorithm speed and accuracy.