Hi everyone. I would like to know if there is anyone on this forum who has done any serious work on AI or machine learning or anything related to the field. I will also like u to tell us the kind of work u did on it. Even if it is a simple project.

Phew!!! Cricket chirps have never been louder

How about some work on Digital Signal Processing? Neural Networks and stuff? Anyone?

In nigeria here, we dont take science too seriously, we are alway ready to buy already made tech from other countries. Right now there is no immediate hope for those interest in science and tech.
But you join us at facebook.com/sciencefortress. It's a relative new science community hopefully with time we all we benefit from each other.

...and this information is just for discussion (nothing practical) or what exactly?

logica: ...and this information is just for discussion (nothing practical) or what exactly?

Ummmm. I said any work done on it. So it definitely ought to be practical. I have a friend who is doing a school project and his supervisor wants him to do something along the line. So he has asked for my help. And I am here seeking u guys help. Kapisce?

Did a couple of java projects on Neural Networks some years back... Will probably need to review some books to get fit again

AOladipupo: Did a couple of java projects on Neural Networks some years back... Will probably need to review some books to get fit again

U are what I have been looking for. What exactly did the ANN do? And how easy/hard was it to accomplish?

Anyone here ever worked on signal processing, more specifically Matched filters? Please I need assistance

PhenomenonVFX:

U are what I have been looking for. What exactly did the ANN do? And how easy/hard was it to accomplish?

Prediction based on data initially fed into it.

AOladipupo:

Prediction based on data initially fed into it.

Thanks bro. But could u be much more specific? U could save a life or two.

@op does game programming fall under this category?

Well, I wouldnt say "fall under" exactly. I would see it more like "intersect" or "overlap". To be a serious game programmer for some adversary games for instance, u may need to know some artificial intelligence especially some game and decision theory and evolutionary models. It is a bit complicated.
Say for instance u want to write a chess program that can play against a human, u may need to code it with basic chess rules and then give it a sort of reward payment system for the having the least loses possible. So the aim to the game will be to minize its losses. And u could also make it learn new moves the human player uses to maximise his own gain and thereby incorporate those moves into its game plan. There are also some games that dont know the rules until the game starts and have to learn d rules unsupervised.
So basically game programming is more like coding the physics or set of rules while AI will help the computer or intelligent agent in the game adapt itself to new situations within the confines of that set of rules.

*clears throat, beams smile n suddenly sets a serious face* I did a project way back on geno-fuzzy systems (genetic algorithm n fuzzy logic), closely related to artificial neural networks. Am rusty now though bt wouldn't mind helping out where I can. So what seek ye?

I never really had much love for genetic algorithms. But fuzzy logic is considerable.
So Sir Crisnite, pray tell, what exactly did u do on it. And how god was the performance?

PhenomenonVFX: I never really had much love for genetic algorithms. But fuzzy logic is considerable.
So Sir Crisnite, pray tell, what exactly did u do on it. And how god was the performance?

Mehn, I fell in love with GA dt period bt FL, well as d name goes was just fuzzy bt scaled through. My work was on an 'expert' system dt can take the symptoms of a patient(plus d degree of d symptoms) n diagnose d patient for an ailment. Mine was focused on hrt diseases.

It wasn't an easy task esp tryin to fuzzify d non-numerical data like feelings n moods into usable figures bt 1s done it was d coding dt gave me headache.....at d end it workd fine wit d little parameters I imputed.

@Crisnite
Interesting stuff u got there. I am still doing some consultation here. I will still like to hear more from u. So I will let u know as soon as possible.

I think I know one or two in this regards...
Theres A HELL of statistics involved

If you still need help, do let me know...Ive done some work using ANN, Bayes and D-S.
If I can help, Il def render my services

proudlysports: If you still need help, do let me know...Ive done some work using ANN, Bayes and D-S.
If I can help, Il def render my services

Aye aye, captain. Your help will be much appreciated. Please could u tell what exactly u did with it and how good the performance was. Also, what is D-S?

D-S = Demspter Shafer...its a generalized Bayes technique used for Data fusion.

Basically, we built a robot for detection of a wide range of substances.

On the performance, we achieved only about half of overall objectives i.e. building the robot, detection of substances.