Neural Net Predictions

Update time, here is the original post regarding my research on a neural net predictor. Lets see how it did:

  • IBM 105/109/110+          102.36 (X)
  • GOOG 420/425/450+     409.50 (X)
  • QCOM 42.5/45/50+        44.80 (C)
  • YHOO 15/16/17               15 (X) E

So 3 wrong 1 correct, combined with 4 correct previously makes it 5 out of 8. That isn’t bad considering given the insignifant sample size and lets keep in mind that previous forecasts were actually ‘easier’ since the market remained relatively flat. As of now we are down about 250 bps on the SPY so we’ve had a large quick move. From my understanding the inputs are limited to OHLCV data, perhaps a solid neural net would incorporate additional inputs which were signaling sell, volatility for example.

I am beginning to experiment with a program RapidMiner, to hopefully create my own NNs and test these ideas.

Advertisements

~ by largecaptrader on July 2, 2009.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: