project checklist

frame the problem

select a performance measure

RMSE:均方误差根
MAE: 平均绝对误差
范数越大对大特征值更有效，会忽略小特征值，但数据正态分布时，RSEM性能更好。

Download and load the data

Take a quick look at the data strucure

data.head()

data.info()
data[‘attribute’].value_counts()
data.describe()
也可以画直方图来了解各个数字型属性的分布
data.hist(bins = 50,figsize=(20,15))

create a test set

random select
from sklearn.model_selection import train_test_split
train_set,test_set = train_test_split(data,test_size = 0.2, random_state = 42)
stratified sampling通过对分组属性进行分层采样划分
from sklearn.model_selection import StratifiedShuffleSplit
spliter = StratifiedShuffleSplit(n_splits = 1,test_size = 0.2,random_state = 42)
for train_index,test_index in spliter.split(data,data[‘category’]):
strat_train_set = data.loc[train_index]
start_test_set = data.loc[test_index]

exploring the data:discover and visualize the data to gain insights

visualizing geographical data
housing.plot(kind = ’scatter’,x= ‘longitude’,y = ‘latitude’ ,alpha = 0.4,s = housing[‘population’]/100,label = ‘population’,c = “median_house_value”,camp = plt.get_cmap(“jet”),colorbar = Ture)
plt.legend()

looking for correlations