Random forest accuracy evaluation

Loss on training data:

In []: 
rf_model.score(X_train, y_train) 
Out[]: 
0.98999999999999999 

Loss on test data:

In []: 
rf_model.score(X_test, y_test) 
Out[]: 
0.90333333333333332 

Cross-validation:

In []:
scores = cross_val_score(rf_model, features, df.label, cv=10) 
np.mean(scores) 
Out[]: 
0.89700000000000002 
In []: 
print("Accuracy: %0.2f (+/- %0.2f)" % (scores.mean(), scores.std() * 2)) 
Accuracy: 0.90 (+/- 0.06) 

Precision and recall:

In []: 
predictions = rf_model.predict(X_test) 
predictions = np.array(map(lambda x: x == 'rabbosaurus', predictions), dtype='int') 
true_labels = np.array(map(lambda x: x == 'rabbosaurus', y_test), dtype='int') 
precision_score(true_labels, predictions) 
Out[]: 
0.9072847682119205 
In []: 
recall_score(true_labels, predictions) 
Out[]: 
0.90131578947368418 

F1-score:

In []: 
f1_score(true_labels, predictions) 
Out[]: 
0.90429042904290435 

Confusion matrix:

In []: 
confusion_matrix(y_test, rf_model.predict(X_test)) 
Out[]: 
array([[134,  14], 
       [ 15, 137]]) 

You export a random forest for the iOS in the same way you do for a decision tree.