https://github.com/lijingpeng
kaggle地址
數據預覽
首先載入數據集
import pandas as pd
import numpy as np
train = pd.read_csv('/Users/frank/Documents/workspace/kaggle/dataset/digit_recognizer/train.csv')
test = pd.read_csv('/Users/frank/Documents/workspace/kaggle/dataset/digit_recognizer/test.csv')
print train.head()
print test.head() label pixel0 pixel1 pixel2 pixel3 pixel4 pixel5 pixel6 pixel7 \
0 1 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0 0
2 1 0 0 0 0 0 0 0 0
3 4 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0 0
pixel8 ... pixel774 pixel775 pixel776 pixel777 pixel778 \
0 0 ... 0 0 0 0 0
1 0 ... 0 0 0 0 0
2 0 ... 0 0 0 0 0
3 0 ... 0 0 0 0 0
4 0 ... 0 0 0 0 0
pixel779 pixel780 pixel781 pixel782 pixel783
0 0 0 0 0 0
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
[5 rows x 785 columns]
pixel0 pixel1 pixel2 pixel3 pixel4 pixel5 pixel6 pixel7 pixel8 \
0 0 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0 0
2 0 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0 0
4 0 0 0 0 0 0 0 0 0
pixel9 ... pixel774 pixel775 pixel776 pixel777 pixel778 \
0 0 ... 0 0 0 0 0
1 0 ... 0 0 0 0 0
2 0 ... 0 0 0 0 0
3 0 ... 0 0 0 0 0
4 0 ... 0 0 0 0 0
pixel779 pixel780 pixel781 pixel782 pixel783
0 0 0 0 0 0
1 0 0 0 0 0
2 0 0 0 0 0
3 0 0 0 0 0
4 0 0 0 0 0
[5 rows x 784 columns]
分離訓練數據和標籤：
train_data = train.values[:,1:]
label = train.ix[:,0]
test_data = test.values降維
from sklearn.decomposition import PCA
from sklearn.svm import SVC
pca = PCA(n_components=0.8, whiten=True)
# pca.fit(train_data)
train_data = pca.fit_transform(train_data)
# pca.fit(test_data)
test_data = pca.transform(test_data)SVM訓練
print('使用SVM進行訓練...')
svc = SVC(kernel='rbf',C=2)
svc.fit(train_data, label)
print('訓練結束.')使用SVM進行訓練...
訓練結束.
print('對測試集進行預測...')
predict = svc.predict(test_data)
print('預測結束.')對測試集進行預測...
預測結束.
保存結果：
pd.DataFrame(
{"ImageId": range(1, len(predict) + 1), "Label": predict}
).to_csv('output.csv', index=False, header=True)
print 'done.'done.