qiujiahao123的博客

import pandasimport numpy as np# Set index_col to False to avoid pandas thinking that the first column is row indexes (it's age).columns = ["age", "workclass", "fnlwgt", "education", "education_num",

#多分类的思想是：将多个类别拆成多个二分类的问题#本实验为三分类问题import pandas as pdimport matplotlib.pyplot as pltcolumns = ["mpg","cylinders","displacement","horsepower","weight","acceleration","model year","origin","car name"]

import numpy as npimport pandas as pdadmission = pd.read_csv("D:\\test\machineLearning\\admissions.csv")print admission.head(2) admission["actul_label"] = admission["admit"]adm

import pandas as pdimport numpy as npfrom sklearn.linear_model import LogisticRegressionnp.random.seed(8)admissions = pd.read_csv("D:\\test\machineLearning\\admissions.csv")admissions["actual_label

true代表你判断的是正确的，positive代表是正力（即是你的目标） false代表你的判断是错误的，negative代表是负力（即不是你的目标） 本例子当中，女生是目标即positive，TP，T代表判断正确，P代表你认为它是一个女生，加起来就是它真的是一个女生。 FP，f代表判断错误，你认为它是一个女生，加起来就是你认为它是一个女生，实际它是一个男生 用于检验positive的有效性

逻辑回归用于分类，可以将任何数据变成0到1之间的数据 import pandas as pdimport numpy as npimport matplotlib.pyplot as pltadmissions = pd.read_csv("D:\\test\machineLearning\\admissions.csv")plt.scatter(admissions["gpa"],admi

线性回归公式: import pandas as pdimport matplotlib.pyplot as pltcolumns = ["mpg","cylinders","displacement","horsepower","weight","acceleration","model year","origin","car name"]#由于原使数据只有数据，因此columns是为他们增

找一条“胖”的决策边界的前提时点不能分错，这条胖的边界我们称为margin 此处distance是图中的h，即点到决策边界的计算公式 优化后的最终公式是下图

1.第一步:前向传播，X经过计算得到L值 第二步:反向传播，根据得到的L值，分析各个W所起到的左右 第三步:调整各个W，使得L值不断下降B和L后面存在一个激活函数，使得神经网络具备非线性功能 正则化项用于解决棍和现象

中间的矩阵S即是我们所关心，它的对角线上的每一个值，越大，说明这一类元素越重要，同时我们也可以取一个r乘以r的近似值，来得到我们所关心的，对最终结果有重要影响的类别。例如，我想直到最重要的2项指标，那么我可以使s成为2行2列的矩阵，u使m*2的矩阵，V是n行2列的矩阵 举例如下: 通过S将U，V控制成2列，这样方便我们在二维的坐标系中展示

#无监督：数据没有明确说明属于哪一类，无需去训练模型import pandas as pdvotes = pd.read_csv("D:\\test\machineLearning\\114_congress.csv")print(votes["party"].value_counts())print votes.mean()R 54D 44I 2Name: par

通过判断彼此间的距离来实现聚类#无监督：数据没有明确说明属于哪一类，无需去训练模型import pandas as pdvotes = pd.read_csv("D:\\test\machineLearning\\114_congress.csv")print(votes["party"].value_counts())print votes.mean()R 54D 44I

#随机森林存在2重随机性：1样本的随机性 2特征的随机性import pandascolumns = ["age", "workclass", "fnlwgt", "education", "education_num", "marital_status", "occupation", "relationship", "race", "sex", "capital_gain

import pandascolumns = ["age", "workclass", "fnlwgt", "education", "education_num", "marital_status", "occupation", "relationship", "race", "sex", "capital_gain", "capital_loss", "hours_per

以下是代码实现#coding:utf-8'''Created on 2017-03-17@author: qiujiahao'''import numpy as npimport operatorfrom audioop import reversedef createDataSet(): #以矩阵的格式创造四个数据 group = np.array([[1.0,1.1