Hi Ron, and all --
I have done a lot of work with neural networks. Neural networks are used for supervised learning. Thinking of the data as a matrix or spreadsheet, there is a row for each price bar; a column for each predictor variable (pattern or indicator); a single column for the target you wish to predict. Rows must be independent, implying that any references to previous values must be precomputed, each then stored in a column of its own. Use AmiBroker to compute the indicators and their lagged values, and the target, then write / export that to a csv file where the NN program can read it.
The NN processing will create the best model it can fitting the predictor variables to the target. Best implies ranking alternative models using an objective function. The NN program you use may give some choices of objective functions. Use the same criteria for selecting the objective function for the NN that you would use / did use when selecting the objective function for the AmiBroker implementation.
When the fitting is complete, the prediction process is -- form additional rows of predictor variables not previously seen by the NN during training and have the stored NN compute the value for the target for each row.
Pay particular attention to the propensity of NNs to overfit.
Pay particular attention to the propensity of NNs to overfit.
Accurate predictions require that the future resembles the past -- that is, that the relationships are stationary throughout the data periods used to train the model and continuing on for the out-of-sample period.
The vast majority of the time and effort spent in any machine learning model building is in data preparation -- choosing a target, choosing an objective function, choosing raw data, cleaning and transforming raw data into accurate predictor variables. You will need to select some metavariables -- the configuration of the neural network, time period for training, time period for testing.
Which neural network program have you decided to use? What others did you consider? What are the important features of the one you chose?
Expect to spend months of full time effort.
Keep us posted as to your results.
On Fri, Jan 29, 2016 at 11:56 AM, 'Ron J' ronj@san.rr.com [amibroker] <amibroker@yahoogroups.com> wrote:
RegardsRon
RonJaenisch.com
"Success is almost totally dependent upon drive and persistence. The extra energy required to make another effort or try another approach is the secret of winning."
— Denis Waitley: American motivational speaker
All of my emails are for Educational or Entertainment purposes only. If you need professional advice please contact an appropriate professional, none is given via my emails.
__._,_.___
Posted by: Howard B <howardbandy@gmail.com>
| Reply via web post | • | Reply to sender | • | Reply to group | • | Start a New Topic | • | Messages in this topic (4) |
**** IMPORTANT PLEASE READ ****
This group is for the discussion between users only.
This is *NOT* technical support channel.
TO GET TECHNICAL SUPPORT send an e-mail directly to
SUPPORT {at} amibroker.com
TO SUBMIT SUGGESTIONS please use FEEDBACK CENTER at
http://www.amibroker.com/feedback/
(submissions sent via other channels won't be considered)
For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG:
http://www.amibroker.com/devlog/
This group is for the discussion between users only.
This is *NOT* technical support channel.
TO GET TECHNICAL SUPPORT send an e-mail directly to
SUPPORT {at} amibroker.com
TO SUBMIT SUGGESTIONS please use FEEDBACK CENTER at
http://www.amibroker.com/feedback/
(submissions sent via other channels won't be considered)
For NEW RELEASE ANNOUNCEMENTS and other news always check DEVLOG:
http://www.amibroker.com/devlog/
.
__,_._,___
EmoticonEmoticon