Titanic: Machine Learning from Disaster

我使用了逻辑回归模型做的分类,只使用了7个特征,很显然分类效果很差只有43.54%,先附上自己的渣代码,后续优化

import matplotlib.pyplot as pltimport pandas as pdimport numpy as npfrom pydoc import describefrom plot_test import weightdef sigmoid(inX):    inX = inX.astype(float)    return 1.0 / (1.0 + np.exp(-1.0*inX))def gradAscent(dataMat, classLabel):    m, n = dataMat.shape    alpha = 0.001    maxCycles = 500    weights = np.ones((n, 1))#     print 'weight:\n',weights    for k in range(maxCycles):        h = sigmoid(dataMat * weights)        error = classLabel - h        weights = weights + alpha * dataMat.T * error    return weightsdef getData(train, test):    train = pd.read_csv(train, header=0)    m, n = train.shape    train.Age = train.Age.fillna(train.Age.median())    train.loc[train.Sex == 'male', 'Sex'] = 1    train.loc[train.Sex == 'female', 'Sex'] = 0    train.Embarked = train.Embarked.fillna('S')    train.loc[train.Embarked == 'S', 'Embarked'] = 0    train.loc[train.Embarked == 'C', 'Embarked'] = 1    train.loc[train.Embarked == 'Q', 'Embarked'] = 2    label_mat = np.mat([train.Survived]).T    data_mat = np.mat([train.Pclass, train.Sex, train.Age, train.SibSp, train.Parch, train.Fare, train.Embarked]).T    data_mat = np.hstack((np.ones((m,1)), data_mat))    test = pd.read_csv(test, header=0)    m2, n2 = test.shape    test.Age = test.Age.fillna(test.Age.median())    test.loc[test.Sex == 'male', 'Sex'] = 1    test.loc[test.Sex == 'female', 'Sex'] = 0    test.Embarked = test.Embarked.fillna('S')    test.loc[test.Embarked == 'S', 'Embarked'] = 0    test.loc[test.Embarked == 'C', 'Embarked'] = 1    test.loc[test.Embarked == 'Q', 'Embarked'] = 2    print test.Fare[152]    print test.Fare.median()    print test.Pclass    test.Fare = test.Fare.fillna(test.Fare.median())    test_mat = np.mat([test.Pclass, test.Sex, test.Age, test.SibSp, test.Parch, test.Fare, test.Embarked]).T    test_mat = np.hstack((np.ones((m2,1)), test_mat))#     print data_mat.shape#     print test_mat.shape#     print label_mat.shape#     print data_mat[0]#     print test_mat[0]    return data_mat, test_mat, label_matdata, test, label = getData('train.csv', 'test.csv')weights = gradAscent(data, label)count = 0f = open('re.csv', 'w')k = 0for i in range(892, 1310):    print k    t = int(sigmoid(test[k]*weights))    k += 1    temp = str(i)+','+str(t)+'\n'    f.write(temp)f.close()