Logistic 多分类

最新推荐文章于 2024-04-12 14:27:13 发布

Wayne2019

最新推荐文章于 2024-04-12 14:27:13 发布

阅读量975

点赞数

分类专栏：深度学习与机器学习文章标签： python

本文链接：https://blog.csdn.net/Wayne2019/article/details/78859242

版权

深度学习与机器学习专栏收录该内容

23 篇文章 3 订阅

订阅专栏

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Wed Dec 20 14:54:31 2017

@author: wayne

"""

import pandas as pd
import numpy as np
from sklearn import preprocessing
from sklearn.linear_model import LogisticRegression

df = pd.read_excel("data.xlsx")
df = df.dropna()

cat_vars = ['relationship', 'class', 'scenario', 'object']
target = 'violence'
Y = df[target]

for var in cat_vars:
    cat_list = pd.get_dummies(df[var], prefix=var)
    df = df.join(cat_list)

delete_vars = cat_vars + list(target)
df_vars = df.columns.values.tolist()
to_keep = [i for i in df_vars if i not in delete_vars]
X = df[to_keep]



min_max_scaler = preprocessing.MinMaxScaler(feature_range=(0,1))
X_scale = min_max_scaler.fit_transform(X)


clf = LogisticRegression()
clf.fit(X_scale, Y)
coefficients = pd.DataFrame({"Feature":X.columns,"Coefficients0":np.transpose(clf.coef_[0,:])\
                             ,"Coefficients1":np.transpose(clf.coef_[1,:])\
                             ,"Coefficients2":np.transpose(clf.coef_[2,:])})
intercepts = clf.intercept_
    
predicted = clf.predict(X_scale)
accuracy_train = (predicted == Y).mean()
print(accuracy_train)