LinearSVC does not have predict_proba() and predict_log_proba()
implementations, but SVC does. This is because liblinear does not
have calibrated probabilities as an option but libsvm does.
Would it be okay if I add a classifier mixin to core that adds Platt
Scaling into the LinearSVC? I already
Sounds like you're on the right path. Looking at the misclassified
documents and the feature coefficients is a common way to debug a
classifier, especially if you use boolean features.
If you're using a sklearn vectorizer this might be of interest to you:
http://stackoverflow.com/questions/669
It is correct to assume that a positive coefficient contributes positively
to a decision.
However, because the features are interdependent, the raw strength of a
feature isn't always straightforward to interpret. For example, it might
give a big positive coefficient to "Tel" and a similar negative
Hello scikit!
I need some insights into what I am doing.
Currently I am doing a text classifier (2 classes) using unigrams (word
level) and some writing style features. I am using a Logistic Regression
model, with L1 regularization. I have a decent performance (around .70
f-measure) for the given
I confirm Python 3.3 is fully supported and that this is a numpy
related warning. PR to fix such warnings are always welcome.
--
Olivier
--
Managing the Performance of Cloud-Based Applications
Take advantage of what the
