Closed Saurabh7 closed 10 years ago
Nice! I will merge this for now and then explore a bit on my machine. Great with the Boston Housing dataset!
@karlnapf where should I add the dataset file?
Locally i had put it in /data/toy since : DATA_PATH = ROOT_PATH + '/data'
shogun-data repository (thats where the other datasets are)
please put your extensions into the README.md
I like this, it would be cool to have it for other methods.
About parameters, I would completely remove them in fact. Rather do grid-search or similar to learn them from data. The point here is to illustrate what the methods do. You can add a little text box where the optimal parameters are printed maybe? Let me know what you think
Importing the Boston housing dataset doesnt work here. Any ideas?
I like this a lot!
Something similar should exist for classification. In fact, try to put in as many models as you can for both regression and classification. And again, just remove the parameters and learn them from data. Quite cool!
Reference your notebook Add LARS
The housing data is not in shogun-data. Could you send a PR with that?
@karlnapf I have added a demo about the regression models: linear ridge, least squares, kernel ridge to choose from. Also a choice to load the Boston housing dataset and its various attributes to play with. Please have a look..