Update - Better late than never!

OK, so its been a while but i figured it was about time I reported on my progress so far.

We've been back a few weeks now and so far the majority of the time has been taken up by the busy work of getting the framework up and running. It took longer than I expected to re familiarise myself with OGRE and as a result I'm a little bit behind in my plan.

The first problem encountered was getting the new version of OGRE working and creating a window, sounds simple? Of course it wasn't! The new version of OGRE required newer versions of other libraries, which I didn't have. However once this was discovered everything came together fairly quickly.

It is essential that some sort of GUI is Incorporated into the application so that the user can navigate and interact with the environment as they desire. For this reason CEGUI was Incorporated into the framework. This again proved to be far more troublesome that it should have been. While the GUI libraries have now been successfully included into the framework, they are still not functional within the program. It is hoped that this problem with reveal itself soon, but enough time has been spent focusing on this so it has been put onto the back burner for now. I also said that there would be a level of audio in the application so FMOD has been included and a sound manager added so that audio can be added when the time comes. This was comparatively easy (thank god) so didn't take up much time.

The rest of the time has been spent into further research and application of decision trees. I created a simple application where a decision tree is used to establish what someone should do with their spare time. Currently it is hard coded and not generated from a data set. It also requires user input and does not reach a decision by itself. Its a trivial example, but its a start and hopefully I'll be able to build upon this soon.

The next two weeks will be spent building upon the start I've made with the decision trees application. It is my aim to have the ID3 algorithm included by then.
This algorithm builds a decision tree based on a data set. It builds the tree in a way that aims to reach the final decision as quick as possible. It can do this via the use of probability or entropy. I'm not sure which way will work the best so I will attempt to implement both and see which one works better.

I also want to make some head way on the graphical side of things. I hope to have a basic version of the environment and camera included soon, so that I can get some intelligent agents in as soon as possible.