sklearn的初尝试

最近学了一些机器学习的架构和内部实现，在实际做project的时候了解到大家都是使用high-level封装过的python机器学习包。sklearn是一个很著名的包，记录一下我对它的初次尝试。

Load Data

# for some strange python import mechanism, can't say
# import sklearn, and then use sklearn.datasets
from sklearn import datasets 
boston = datasets.load_boston()

boston_X = boston.data
boston_y = boston.target

Model Set Up

from sklearn import linear_model  # for linear regression classifier
# model set up
model = linear_model.LinearRegression()
# training
model.fit(boston_X, boston_y)

Prediction

from sklearn import neighbors  # for knn classifier

# knn classifier set up
knn = neighbors.KNeighborsClassifier()

# do training
knn.fit(X_train, y_train)

pred = knn.predict(test_inst[0].reshape(1, -1))