Hello Aron,
Excellent example of ML in AFL. Thank you for sharing your work.
Can you please elaborate on how the "feature" set is generated/calculated (if I want to try some other data)?
How is the "target" data set generated/calculated? it appears that in your example it is a synthetic/example.
First 58 lines of the target set are equal to 1 and the reminder is all equal to 0.
Does the LogisticRegression function accept more than two feature columns or do we have to adjust the code for that?
What is the interpretation for the "Classification" XYPlot?
- Green = correctly predicted positives?
- Red = correctly predicted negatives?
- Black = misclassified?
Kind Regards
Richard
---In amibroker@yahoogroups.com, <aron@...> wrote :
With Newton's method you do not need to set a learning rate and the algorithm may convergence in a few iterations.
Calculations involve matrix inversion so there might not work in all the cases but when it does its awesome.
fitting a 3-d degree polynomial:
The decision boundary is drawn by hand, just to give an idea.
you can get the exact boundary by solving :
t0 + t1x + t2xy + t3y +
t4x2 + t5x2y + t6xy2 + t7y2 +
t8x3 + t9x3y + t10xy3 + t11y3 = 0
hollow circles represent misclassified instances
Enjoy!
Posted by: richpach2@yahoo.com
| Reply via web post | • | Reply to sender | • | Reply to group | • | Start a New Topic | • | Messages in this topic (14) |
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