Python plot_confusion_matrixの例

コード例 #1

nmse = normalized_mean_squared_error(truthMean_test, truthMean_pred)
mse = mean_squared_error(truthMean_test, truthMean_pred)
evs = explained_variance_score(truthMean_test, truthMean_pred)
r2 = r2_score(truthMean_test, truthMean_pred)
mae = mean_absolute_error(truthMean_test, truthMean_pred)
med_a_e = median_absolute_error(truthMean_test, truthMean_pred)

truthClass_test = [0 if t == 'no-clickbait' else 1 for t in truthClass_test]
truthClass_pred = [0 if t < 0.5 else 1 for t in truthMean_pred]


tn, fp, fn, tp = conf_matrix = confusion_matrix(truthClass_test, truthClass_pred).ravel()
print("(TN, FP, FN, TP) = {}".format((tn, fp, fn, tp)))

compute_diffs(truthClass_test, truthClass_pred, truthMean_test, truthMean_pred)
plot_confusion_matrix(truthClass_test, truthClass_pred, title='Random Forest confusion matrix')

accuracy = accuracy_score(truthClass_test, truthClass_pred)
auc = roc_auc_score(truthClass_test, truthClass_pred)
precision = precision_score(truthClass_test, truthClass_pred)
recall = recall_score(truthClass_test, truthClass_pred)
f1 = f1_score(truthClass_test, truthClass_pred)

print('----------------Regression metrics-------------------------\n')
print("Explained Variance Score is ", evs)
print("Mean Squared Error  is ", mse)
print("Normalized Mean Squared Error  is ", nmse)
print("Mean Absolute Error  is ", mae)
print("Median Absolute Error  is ", med_a_e)
print("R2 score is ", r2)

コード例 #2

evs = explained_variance_score(truthMean_test, truthMean_pred)
r2 = r2_score(truthMean_test, truthMean_pred)
mae = mean_absolute_error(truthMean_test, truthMean_pred)
med_a_e = median_absolute_error(truthMean_test, truthMean_pred)

truthClass_test = [0 if t == 'no-clickbait' else 1 for t in truthClass_test]
truthClass_pred = [0 if t < 0.5 else 1 for t in truthMean_pred]

compute_diffs(truthClass_test, truthClass_pred, truthMean_test, truthMean_pred)

tn, fp, fn, tp = conf_matrix = confusion_matrix(truthClass_test,
                                                truthClass_pred).ravel()
print("(TN, FP, FN, TP) = {}".format((tn, fp, fn, tp)))

plot_confusion_matrix(truthClass_test,
                      truthClass_pred,
                      title='Linear regression confusion matrix')

accuracy = accuracy_score(truthClass_test, truthClass_pred)
auc = roc_auc_score(truthClass_test, truthClass_pred)
precision = precision_score(truthClass_test, truthClass_pred)
recall = recall_score(truthClass_test, truthClass_pred)
f1 = f1_score(truthClass_test, truthClass_pred)

print('----------------Regression metrics-------------------------\n')
print("Explained Variance Score is ", evs)
print("Mean Squared Error  is ", mse)
print("Normalized Mean Squared Error  is ", nmse)
print("Mean Absolute Error  is ", mae)
print("Median Absolute Error  is ", med_a_e)
print("R2 score is ", r2)

コード例 #3

clf.fit(X_train, truthMean_train)
truthMean_pred = clf.predict(X_test)

nmse = normalized_mean_squared_error(truthMean_test, truthMean_pred)
mse = mean_squared_error(truthMean_test, truthMean_pred)
evs = explained_variance_score(truthMean_test, truthMean_pred)
r2 = r2_score(truthMean_test, truthMean_pred)
mae = mean_absolute_error(truthMean_test, truthMean_pred)
med_a_e = median_absolute_error(truthMean_test, truthMean_pred)

truthClass_test = [0 if t == 'no-clickbait' else 1 for t in truthClass_test]
truthClass_pred = [0 if t < 0.5 else 1 for t in truthMean_pred]

compute_diffs(truthClass_test, truthClass_pred, truthMean_test, truthMean_pred)
plot_confusion_matrix(truthClass_test,
                      truthClass_pred,
                      title='Adaboost confusion matrix')

tn, fp, fn, tp = conf_matrix = confusion_matrix(truthClass_test,
                                                truthClass_pred).ravel()
print("(TN, FP, FN, TP) = {}".format((tn, fp, fn, tp)))

accuracy = accuracy_score(truthClass_test, truthClass_pred)
auc = roc_auc_score(truthClass_test, truthClass_pred)
precision = precision_score(truthClass_test, truthClass_pred)
recall = recall_score(truthClass_test, truthClass_pred)
f1 = f1_score(truthClass_test, truthClass_pred)

print('----------------Regression metrics-------------------------\n')
print("Explained Variance Score is ", evs)
print("Mean Squared Error  is ", mse)

コード例 #4

ファイル: final_svr.py プロジェクト: rpytel1/clickbait-challenge

mse = mean_squared_error(truthMean_test, truthMean_pred)
evs = explained_variance_score(truthMean_test, truthMean_pred)
r2 = r2_score(truthMean_test, truthMean_pred)
mae = mean_absolute_error(truthMean_test, truthMean_pred)
med_a_e = median_absolute_error(truthMean_test, truthMean_pred)

truthClass_test = [0 if t == 'no-clickbait' else 1 for t in truthClass_test]
truthClass_pred = [0 if t < 0.5 else 1 for t in truthMean_pred]

tn, fp, fn, tp = conf_matrix = confusion_matrix(truthClass_test,
                                                truthClass_pred).ravel()
print("(TN, FP, FN, TP) = {}".format((tn, fp, fn, tp)))

compute_diffs(truthClass_test, truthClass_pred, truthMean_test, truthMean_pred)
plot_confusion_matrix(truthClass_test,
                      truthClass_pred,
                      title='SVR confusion matrix')

accuracy = accuracy_score(truthClass_test, truthClass_pred)
auc = roc_auc_score(truthClass_test, truthClass_pred)
precision = precision_score(truthClass_test, truthClass_pred)
recall = recall_score(truthClass_test, truthClass_pred)
f1 = f1_score(truthClass_test, truthClass_pred)

print('----------------Regression metrics-------------------------\n')
print("Explained Variance Score is ", evs)
print("Mean Squared Error  is ", mse)
print("Normalized Mean Squared Error  is ", nmse)
print("Mean Absolute Error  is ", mae)
print("Median Absolute Error  is ", med_a_e)
print("R2 score is ", r2)