In [6]:

    

import numpy as np
from sklearn import linear_model
from sklearn.ensemble import GradientBoostingRegressor
from sklearn.model_selection import KFold
from sklearn.metrics import mean_squared_error, mean_absolute_error
from sklearn.model_selection import cross_val_score
#####################################################################
# DataSet
#####################################################################
train_file = open('../../0.part.tokens.LTR_FT', 'rb')
character = {}
for line in train_file:
    terms = line.split('\t')
    key = terms[0] + ' ' + terms[1]
    if not key in character:
        character[key] = []
    character[key].append(terms[2])
train_file.close()
samples = character.values()


    

In [14]:

    

def sample2Xy(samples, indexes):
    X = []
    y = []
    for index in indexes:
        for des in samples[index]:
            terms = des.split(' ')
            X.append([float(x) for x in terms[1:]])
            y.append(float(terms[0]))
    X = np.array(X, dtype=np.float32)
    y = np.array(y, dtype=np.float32)
    return X, y
kf = KFold(n_splits = 5, shuffle = True)
mse_linear_model = []
mae_linear_model = []
mse_tree_model = []
mae_tree_model = []
for train, test in kf.split(samples):
    X_train, y_train = sample2Xy(samples, train)
    X_test, y_test = sample2Xy(samples, test)
    # Linear Regression
    clf = linear_model.LinearRegression()
    clf.fit(X_train, y_train)
    y_pred = clf.predict(X_test)
    mse_linear_model.append(mean_squared_error(y_test, y_pred))
    mae_linear_model.append(mean_absolute_error(y_test, y_pred))
    # Gradient Boosting Regressor
    clf = GradientBoostingRegressor(learning_rate=0.05, random_state=1)
    clf.fit(X_train, y_train)
    y_pred = clf.predict(X_test)
    mse_tree_model.append(mean_squared_error(y_test, y_pred))
    mae_tree_model.append(mean_absolute_error(y_test, y_pred))
print 'Linear Regression MSE ' + str(abs(np.mean(mse_linear_model)))
print 'Linear Regression MAE ' + str(abs(np.mean(mae_linear_model)))
print 'Gradient Boosting Regressor MSE ' + str(abs(np.mean(mse_tree_model)))
print 'Gradient Boosting Regressor MAE ' + str(abs(np.mean(mae_tree_model)))


    

    




Linear Regression MSE 0.0421227189918
Linear Regression MAE 0.164648788162
Gradient Boosting Regressor MSE 0.0323571346827
Gradient Boosting Regressor MAE 0.141880088677

In [65]:

    

i = 495
print y_test[i: i + 10]
print y_pred[i: i + 10]


    

    




[ 0.44574937  0.32700461  1.          1.          0.48753181  1.          1.
  0.47809133  1.          0.88545078]
[ 0.8824008   0.48452059  0.90852744  0.84650243  0.56198991  0.94819097
  0.93412086  0.6356297   0.99729832  0.6399198 ]