Skip to content

Given a dataset containing car attributes, use MLP and RBF networks to predict the Miles per Gallon consumption

Notifications You must be signed in to change notification settings

paulrox/MPG_Prediction

Repository files navigation

Miles per Gallon Prediction

Given a dataset containing the features (weigth, country, year, etc.) of several cars, compare the architectures of different Multi-Layer Perceptron and Radial Basis Function networks built with MATLAB and choose the one which gives the smallest error.

Features Selection

Not all the features in the dataset are useful for the training of the MLPs and for this reason a subset of them have been choosen by using a Genetic Algorithm able compare the correlation of different subset of features with the MPG.

Evaluation of the Networks

After choosing the features, a set of possible architectures for RBF and MLP networks have been evaluated by training them using the choosen set of features.

The whole process (features selection + network evaluation) can be executed using the script main.m.

Authors

About

Given a dataset containing car attributes, use MLP and RBF networks to predict the Miles per Gallon consumption

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages