mdl_estimation

gistfile1.py

# quality of classification
from sklearn.metrics import accuracy_score
y_true = test['numeric_status']
y_pred = result
print "Predicted accuracy: %f" % accuracy_score(y_true, y_pred)

# let's also calculate "baseline accuracy" - when we just say "yes" every time (most common outcome)
baseline = [1] * len(y_pred)
print "Baseline accuracy: %f" % accuracy_score(y_true, baseline)