ML
文章平均质量分 87
爱跑步的george
北京邮电大学
展开
-
ML3: sklearn 线性回归
import numpy as np import sklearn.linear_model as lm import sklearn.metrics as sm import matplotlib.pyplot as mp X,Y = [],[]with open('dd.txt','r') as f: for line in f.readlines(): ...原创 2018-06-05 15:30:20 · 174 阅读 · 0 评论 -
ML13: 随机森林预测
from __future__ import unicode_literalsimport csvimport numpy as npimport sklearn.utils as suimport sklearn.ensemble as seimport sklearn.metrics as smimport matplotlib.pyplot as mpwith open('.....原创 2018-06-06 09:59:06 · 704 阅读 · 0 评论 -
ML14: 简单分类器
对于可以人工方法找到的分类依据就没有必要使用数学模型完成分类,换言之自己编写的代码逻辑就已经体现分类模型的效果。 # -*- coding: utf-8 -*-from __future__ import unicode_literalsimport numpy as npimport matplotlib.pyplot as mpx = np.array([ [3, 1...原创 2018-06-06 12:34:41 · 393 阅读 · 0 评论 -
ML15: 机器学习2
十、逻辑分类器1.预测函数g(z) = 1 / (1+e^-z) z = t0 + t1x1 + t2x2 + … + tnxn h(x1, …, xn) = 1 / (1+e^-(t0 + t1x1 + t2x2 + … + tnxn)) J(t0, …, tn) = sigma((h(x1, …, xn) - y)^2) / 2m t0, …, tn ? -> J(t...原创 2018-06-06 14:53:24 · 477 阅读 · 2 评论 -
ML16:逻辑回归
1.预测函数 g(z) = 1 / (1+e^-z) z = t0 + t1x1 + t2x2 + … + tnxn h(x1, …, xn) = 1 / (1+e^-(t0 + t1x1 + t2x2 + … + tnxn)) J(t0, …, tn) = sigma((h(x1, …, xn) - y)^2) / 2m t0, …, tn ? -> J(t0, …, tn) m...原创 2018-06-06 15:00:12 · 259 阅读 · 0 评论 -
ML17:逻辑回归多分类
# -*- coding: utf-8 -*-from __future__ import unicode_literalsimport numpy as npimport sklearn.linear_model as lmimport matplotlib.pyplot as mpx = np.array([ [4, 7], [3.5, 8], [...原创 2018-06-06 15:32:09 · 179 阅读 · 0 评论 -
ML18: 朴素贝叶斯多分类
# -*- coding: utf-8 -*-from __future__ import unicode_literalsimport numpy as npimport sklearn.naive_bayes as nbimport matplotlib.pyplot as mpx, y = [], []with open('/Users/youkechaung/Desktop/算...原创 2018-06-06 15:40:50 · 2289 阅读 · 0 评论 -
ML19: sklearn 划分测试集+交叉验证+混淆矩阵+随机森林汽车品质评估案例
19划分测试集# -*- coding: utf-8 -*-from __future__ import unicode_literalsimport numpy as npimport sklearn.model_selection as msimport sklearn.naive_bayes as nbimport matplotlib.pyplot as mpx, y = ...原创 2018-06-07 08:39:16 · 6126 阅读 · 0 评论 -
ML12:特征重要性 选取
from __future__ import unicode_literalsimport numpy as npimport sklearn.datasets as sdimport sklearn.utils as suimport sklearn.tree as stimport sklearn.ensemble as seimport matplotlib.pyplot as ...原创 2018-06-06 09:45:04 · 882 阅读 · 0 评论 -
ML11: 决策树DecisionTreeRegressor 自适应增强决策树AdaBoostRegressor 波士顿房价预测
import sklearn.datasets as sd import sklearn.utils as su import sklearn.tree as st import sklearn.ensemble as se import sklearn.metrics as sm housing = sd.load_boston()x,y = su.shuffle(housing.d...原创 2018-06-06 09:05:06 · 685 阅读 · 0 评论 -
ML4: sklearn np.lstsq MinMaxScaler
from __future__ import unicode_literalsimport sklearn.preprocession as sp raw_samples = np.array([[3,-1.5,2,-5.4],[0.4.-0.3,2.1],[1,3.3,-1.9,-4.3] ])print(raw_samples)print(raw_samples.min(...原创 2018-06-05 15:57:54 · 390 阅读 · 0 评论 -
ML5: sklearn正则化
from __future__ import unicode_literalsimport sklearn.preprocession as sp raw_samples = np.array([[3,-1.5,2,-5.4],[0.4.-0.3,2.1],[1,3.3,-1.9,-4.3] ])print(raw_samples)print(abs(raw_samples)...原创 2018-06-05 16:03:50 · 527 阅读 · 0 评论 -
ML6: Sklearn OneHotencoding
from __future__ import unicode_literalsimport sklearn.preprocession as sp raw_samples = np.array([[3,-1.5,2,-5.4],[0.4.-0.3,2.1],[1,3.3,-1.9,-4.3] ])print(raw_samples)code_tables = []for ...原创 2018-06-05 17:04:39 · 138 阅读 · 0 评论 -
ML7: Sklearn scale mean std
# -*- coding: utf-8 -*-from __future__ import unicode_literalsimport numpy as npimport sklearn.preprocessing as spraw_samples = np.array([ [3, -1.5, 2, -5.4], [0, 4, -0.3, 2.1], ...原创 2018-06-05 17:15:21 · 224 阅读 · 0 评论 -
ML1: Binarizer.py
# -*- coding: utf-8 -*-from __future__ import unicode_literalsimport numpy as npimport sklearn.preprocessing as spraw_samples = np.array([ [3, -1.5, 2, -5.4], [0, 4, -0.3, 2.1], ...原创 2018-06-05 15:12:08 · 112 阅读 · 0 评论 -
ML2: Sklearn.preprocessiong LabelEncoder()
from __future__ import unicode_literalsimport numpy as npimport sklearn.preprocessing as spraw_samples = np.array([ 'audi', 'ford', 'audi', 'toyota', 'ford', 'bmw', 'toyota', 'ford', 'audi'...原创 2018-06-05 15:15:10 · 405 阅读 · 0 评论 -
ML8: 机器学习
机器学习一、机器学习的过程特征选择->数据采集->数据清洗->数据划分->模型选择->模型训练->模型测试->模型验证->模型存储->模型应用二、数据预处理1.均值移除将每个特征的平均值设置为0,将标准差设置为1。 sklearn.preprocessing.scale(原始样本原创 2018-06-05 22:48:20 · 411 阅读 · 0 评论 -
ML9: 线性回归
from __future__ import unicode_literalsimport pickleimport numpy as np import sklearn.linear_model as lm import sklearn.metrics as sm import matplotlib.pyplot as mp #读取训练集x,y = [],[]with open...原创 2018-06-05 22:49:09 · 231 阅读 · 0 评论 -
ML10:多项式回归
from __future__ import unicode_literalsimport pickleimport numpy as np import sklearn.linear_model as lm import sklearn.metrics as sm import matplotlib.pyplot as mp import sklearn.pipeline as pl...原创 2018-06-06 08:55:57 · 146 阅读 · 0 评论 -
什么是先验概率、后验概率和贝叶斯公式
验前概率就是通常说的概率,验后概率是一种条件概率,但条件概率不一定是验后概率。贝叶斯公式是由验前概率求验后概率的公式。 举一个简单的例子:一口袋里有3只红球、2只白球,采用不放回方式摸取,求: ⑴ 第一次摸到红球(记作A)的概率; ⑵ 第二次摸到红球(记作B)的概率; ⑶ 已知第二次摸到了红球,求第一次摸到的是红球的概率。 解:⑴ P(A)=3/5,这就是验前概率; ⑵ P(B)=P(...原创 2018-06-26 20:55:17 · 3127 阅读 · 0 评论