Trading using Machine Learning

Hi Friends,

I am starting this thread to kick start ideas that using Machine Learning(Main Focus) and also other advance systems for Trading. I have created a couple of strategies that use clustering and advanced time series analysis (the results of which i will post soon).

Why Machine learning?
Technical analysis based trading has been in the game for long time now and many indicators fail at certain positions. Plan is to overcome and build filters that sit above standard technical indicators and filter to give more accurate profit calls.
Hi all,

I am trying to make a consolidated report on the strategy. I classify profits and losses from historical data based on fixed strategy and then apply clustering techniques to classify them.

I have following question for presenting the report. Answers to these will keep my report precise to the interest of the user and the community.

1. We used filtering over MAC(34 period and 89 period). What kind of results should i present? Can any one share sample?
2. What is % return is considered as good strategy
3. % of profit vs loss trades?
4. Frequency of calls.