regression
regression copied to clipboard
Regression algorithms
Using `new ExponentialRegression([0, 1, 2] [-1, -2, -5], 2)` results in `NaN` returned prediction.
Or put documentation directly in the readme.
https://en.wikipedia.org/wiki/Simple_linear_regression#Linear_regression_without_the_intercept_term
Hi @andcastillo. I am doing some cleanup on this package, and splitting it into multiple projects for more visibility on npm. I wonder about the rationale for the `PotentialRegression` class....
We have currently a demonstration of the tool on: http://www.cheminfo.org/ML/Regression/Simple_Linear_Regression/index.html?viewURL=https%3A%2F%2Fcouch.cheminfo.org%2Fcheminfo-public%2F420cf9177e43f89292d5ca966213ab06%2Fview.json%3Frev%3D1-6a4f3bb721b7976d9ff3e346a2fe2a07 But it is quite old and should be updated to show all the possibilities of the regression package. I would...
Would be nice that in the model you directly return as well the error Which error should be calculated ? Currently in the iamage project there is this: ``` var...
The branch ridge-regression is kernel ridge-regression Implement : http://www.mathworks.com/matlabcentral/fileexchange/31636-2d-polynomial-fitting-with-svd
See: https://github.com/mrocklin/multipolyfit https://github.com/ahmetcecen/MultiPolyRegress-MatlabCentral http://www.researchgate.net/publication/264425037_Multivariate_Polynomial_Regression_in_Data_Mining_Methodology_Problems_and_Solutions