学习人工智能-用python进行训练步骤，DecisionTreeClassifier模型，Graphviz Interactive Preview 预览.dot文件

用python进行训练步骤：

1. 导入数据 import the data
2. 清理数据 clean the data
3. 把数据分为训练和验证部分 split the data into training and test sets
4. 创建一个模型 create a model
5. 训练模型 train the model
6. 创建预测 make predictions
7. 打分评估 score and evaluate.

以DecisionTreeClassifier为例：

直接把数据灌入模型，然后预测

import pandas as pd
from sklearn.tree import DecisionTreeClassifier

music_data = pd.read_csv('music.csv')
x =  music_data.drop(columns=['genre'])
y = music_data['genre']

model  = DecisionTreeClassifier()
model.fit(x,y)

predictions  = model.predict([[21,1],[22,0]])
print(predictions)

把数据分为训练数据和验证数据，再训练，之后打分。

import pandas as pd
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

music_data = pd.read_csv('music.csv')
x = music_data.drop(columns=['genre'])
y = music_data['genre']
x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.1)


model = DecisionTreeClassifier()
model.fit(x_train, y_train)

predictions  = model.predict(x_test)


score_results = accuracy_score(y_test, predictions)
print(score_results)

训练后把训练的结果存入文件，以后可以直接载入文件进行预测。

import pandas as pd
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
import joblib


music_data = pd.read_csv('music.csv')
x = music_data.drop(columns=['genre'])
y = music_data['genre']
x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.1)
model = DecisionTreeClassifier()
model.fit(x_train, y_train)

#把训练完成的结果输出到文件。
joblib.dump(model, 'music-rec.joblib')

#直接载入 训练完成的文件，用来预测。
model = joblib.load('music-rec.joblib')
predictions = model.predict([[21,1],[22,0]])
print(predictions)

生成.dot文件，用VS code生成预览，查看训练结束的逻辑。

import pandas as pd
from sklearn.tree import DecisionTreeClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
import joblib

from sklearn import tree


music_data = pd.read_csv('music.csv')
x = music_data.drop(columns=['genre'])
y = music_data['genre']
x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.1)

model = DecisionTreeClassifier()
model.fit(x_train, y_train)

tree.export_graphviz(model,
                     out_file='music-plot.dot',
                     feature_names=['age','gender'],
                     class_names=sorted(y_train.unique()),
                     label='all',
                     rounded=True,
                     filled=True
                     )

在VS code里面安装插件Graphviz Interactive Preview， 预览.dot文件，查看逻辑。