my final year project

toppzy · #1 (**permalink**) 06-19-2008, 01:34 PM

I'm a final year student of computer science i want my project area to be related to forex (Specifically Expert advisors). i;m really thinking of a suitalble topic can anybody suggest one for me?
'lf you have any idea pls drop it here or PM me
thanks

IN10TION · #2 (**permalink**) 06-19-2008, 01:48 PM

Call it: " the crusade for the Holy Grail of FOREX, automated moneymakers "

just joking

... or is this serious business!!
good luck

Quote:

Originally Posted by toppzy

I'm a final year student of computer science i want my project area to be related to forex (Specifically Expert advisors). i;m really thinking of a suitalble topic can anybody suggest one for me?
'lf you have any idea pls drop it here or PM me
thanks

luxinterior · #3 (**permalink**) 06-19-2008, 01:55 PM

Where are you in your development? Do you know how to program an EA or are you just talking about the "concept" of EA's?

Lux

4xCoder · #4 (**permalink**) 06-19-2008, 09:57 PM

Here are some suggestions. They might be more suited for a masters thesis as they are very challenging. They are some problems I've been looking at.

Implement a GMDH (GMDH method for forecasting algorithms, fuzzy models analysis) that forecasts a currency pair.

Build a library on top of a functional programming languge like Haskell or F# (http://research.microsoft.com/fsharp/fsharp.aspx) for building EAs in that langugage.

Implement a library for building trading systems using Fuzzy Logic.

Implement an LSTM recurrent network (RECURRENT NEURAL NETWORKS) for forcasting a currency pair.

toppzy · #5 (**permalink**) 06-22-2008, 09:11 PM

Quote:

Originally Posted by luxinterior

Where are you in your development? Do you know how to program an EA or are you just talking about the "concept" of EA's?

Lux

thanks for your reply
i'm not too good with EA's thought but i'm taking my time to learn more about it .
i have just 6 months to learn and come up with my projects

toppzy · #6 (**permalink**) 06-22-2008, 09:15 PM

Quote:

Originally Posted by 4xCoder

Here are some suggestions. They might be more suited for a masters thesis as they are very challenging. They are some problems I've been looking at.

Implement a GMDH (GMDH method for forecasting algorithms, fuzzy models analysis) that forecasts a currency pair.

Build a library on top of a functional programming languge like Haskell or F# (http://research.microsoft.com/fsharp/fsharp.aspx) for building EAs in that langugage.

Implement a library for building trading systems using Fuzzy Logic.

Implement an LSTM recurrent network (RECURRENT NEURAL NETWORKS) for forcasting a currency pair.

whao this is a detailed one thanks soo much.
that gives me something to really think about. 4xCoder i think i'll want to keep in touch with you on this
thanks

TheRumpledOne · #7 (**permalink**) 06-23-2008, 03:59 AM

Why is it that 90% of traders fail instead of 50%?

Wouldn't it be normal to think that 50% would buy when the other 50% would sell? So shouldn't 50% of traders win?

Now that's a riddle to solve!

mikkom · #8 (**permalink**) 06-23-2008, 09:27 AM

Quote:

Originally Posted by 4xCoder

Implement a GMDH (GMDH method for forecasting algorithms, fuzzy models analysis) that forecasts a currency pair.

I just wanted to say thanks for the link. Great stuff.

Other interesting model could be dynamic bayesian network, specifically Hidden Markov model

4xCoder · #9 (**permalink**) 06-23-2008, 03:38 PM

Quote:

Originally Posted by toppzy

whao this is a detailed one thanks soo much.
that gives me something to really think about. 4xCoder i think i'll want to keep in touch with you on this
thanks

np. I'm interested to hear if you choose one of these projects. Keep me informed via PM if you wish. Or my email is 4xcoder [at] gmail [dot] com
Good luck

jbfx · #10 (**permalink**) 06-24-2008, 12:48 AM

Quote:

Originally Posted by TheRumpledOne

Why is it that 90% of traders fail instead of 50%?

Wouldn't it be normal to think that 50% would buy when the other 50% would sell? So shouldn't 50% of traders win?

Now that's a riddle to solve!

Working on something to solve this. The results will likely show the power of the broker's spread and fear at the core of the riddle.