Intel-ML-Week2

#Supervised:   Data points have known outcome
#Unsupervised  Daya points have unknown outcome

#Regression Outcome is continuous(numerical)  #结果是连续的（数值）
#Classification Outcome is a category   #结果是一个分类

#Supervised learning overview
#Data with answer + Model -> Fit -> Model
#Data without answers + Model -> Predict -> Predicted answers

#Regression:Numerical answers
#Movie data with revenue + Model -> Fit -> Model
#Movie data(unkown revenue(收入)) + Model -> Predict -> Predicted revenue

#Classification:categorical(分类) answers
#Labeled data + Model -> Fit -> Model
#Unlabeled data + Model -> Predict -> Labels

#Target:Predicted category or value of the data(column to predict)
#Feature:Properties of the data used for prediction(non-target columns)
#Label:the target value for a single data point

#What is need for classification?
#Model data with:
#    Features that can be quantitated (可以量化的特征)
#    Labels that are known
#Method to measure similarity(相似性度量方法)

#comparison of feature scaling methods
#Standard Scaler:Mean center data and scale to unit variance (平均中心数据和规模单位方差)
#Minimum-Maximum Scaler:Scale data to fixed range(usually 0-1)(将数据缩放到固定范围)
#Maximum Absolute Value Scaler:Scale maximum absolute value(最大绝对值东标)

 

#Import the class containing the scaling method 
from sklearn.preprocessing import StandardScaler
#Create an instance(实例) of the class
StdSc = StandardScaler()
#Fit the scaling parameters and then transform the data (拟合缩放参数，然后转换数据)
StdSc = StdSc.fit(x_data)
x_scaled = KNN.transform(x_data)
#Other scaling methods exist:MinMaxScaler,MaxAbsScaler 

 

#Charecteristics of a KNN model
#Fast to creat model because it simply stores data(快速创建模型，因为只存储数据)
#Slow to predict because many distance calculations(预测缓慢，因为有许多距离需要计算)
#Can require lots of memory if data set is large(如果数据量大，可能需要大量内存)

#K Nearest Neighbors:The Syntax
#Import the class containing the classification method
from sklearn.neighbors import KNeighborsClassifier
#Create an instance of the class
KNN = KNeighborsClassifier(n_neighbors=3)
#Fit the instance on the data and then predict the expected value(在数据上拟合实例，然后预测预期值)
KNN = KNN.fit(x_data, y_data)
y_predict = KNN.predict(x_date)
#Regression can be done with KneighborsRegressor