Last weeks were quite hectic for me... Therefore I haven't really had time to progress with any of the Coursera courses at the moment. I will try to get back to those tomorrow, so you can expect some posts coming up soon.
For now I will write about details of my first AI project. From the previous post you know that I had to train a Neural Network in Neuroph to control a car in TORCS application.
A little bit more information about TORCS:
TORCS is open application used for championships every year. It stands for The Open Racing Car Simulator. Each year people compete against each other to come up with the best controller they can. This is also used for research purposes (mainly by PhD students but it's not limited to them).
To control a car via code TORCS has a number of variables used- those are described in TORCS official documentation. For my project I mainly used steering values as well as sensor values that has given distance of the car from the track.
A bit about Neuroph:
Neuroph is an open source framework that runs on top of NetBeans and is coded with Java. It is usefult for creating Neural Networks via GUI system and then testing them (depending what algorithms are used training data may be required as well). Then saved Neural network can be used again through Java code if using correct libraries.
More coming soon! :)
No comments:
Post a Comment