Machine Learning Voting Committee

Hi,

I'm new to Zorro, but have some years experience in software. I write in Ruby, typically, and have done a little R. My C is rather rusty. I've been through the workshops and enjoyed them a lot.

I haven't traded real money yet, but have been interested in trading for some time. I've done statistical analysis on forex data using both Ruby and R in an attempt to find profitable patterns. I didn't find any good enough to win on.

I'm interested in extending workshop 7 to train several adviseLong and adviseShort calls so that I can use the results in a vote. e.g. train a pattern and a perceptron then only trade of both agree.

Can anyone provide guidance on how I might structure that?

For this you need several algos. Per algo you have one adviseLong and one adviseShort call. Collect the votes from the algos and evaluate them for a trading decision in a final algo.