数据挖掘建模（1）分类与预测

痛定思痛。 2022-05-21 06:09 253阅读 0赞

经过数据探索和数据预处理， 得到了可以直接建模的数据。 根据挖掘目标， 和数据形式可以建立分类与预测、聚类分析、关联规则、时序模型和偏差检测等模型， 帮助企业提取数据中蕴含的商业价值， 提高企业的竞争力。

### 分类与预测： ###

分类模型的实现步骤：

![70][]

常用的分类与预测算法：

![70 1][]

可参考：

[逻辑回归（Logistic Regression）分类算法][Logistic Regression]

[决策树（Decision Tree）分类算法][Decision Tree]

[KNN——K最近邻算法][KNN_K]

朴素贝叶斯分类

[SVM支持向量机][SVM]

[集成学习][Link 1]

人工神经网络

朴素贝叶斯：

设![\\Omega][Omega]是实验E 的样本空间，A为E 的事件，![B\_1, B\_2\\cdots B\_n为\\Omega][B_1_ B_2_cdots B_n_Omega]为![\\Omega][Omega] 的一个划分

> 全概率公式为：
> 
> ![A=A\\Omega =A\\cap (B\_1\\cup B\_2\\cup\\cdots B\_n)\\\\ =AB\_1 \\cup AB\_2 \\cup \\cdots AB\_n\\\\ P(A)=P(B\_1)P(A/B\_1)\\cdotsP(B\_n)P(A/B\_n)][A_A_Omega _A_cap _B_1_cup B_2_cup_cdots B_n_ _AB_1 _cup AB_2 _cup _cdots AB_n_ P_A_P_B_1_P_A_B_1_cdotsP_B_n_P_A_B_n]

![8ed5a02ba5e9856a561260f2_pn_8_o_jpg_6_md5sum_70f9cbdaa8e739f584a4e83b46b54fd3_sign_d885d382d7_png_44692-54685_jpg_524661-612874][]

![8ed5a02ba5e9856a561260f2_pn_11_o_jpg_6_md5sum_70f9cbdaa8e739f584a4e83b46b54fd3_sign_d885d382d7_png_63267-65522_jpg_818745-898101][]

![8ed5a02ba5e9856a561260f2_pn_12_o_jpg_6_md5sum_70f9cbdaa8e739f584a4e83b46b54fd3_sign_d885d382d7_png_65523-65765_jpg_898102-978067][]

[有个朴素贝叶斯决策的例子][Link 2]

[70]: /images/20220521/b6b1518f725b48e58dbbc76192d4d1a4.png
[70 1]: /images/20220521/25c8a425defe4dfba1ca135bed33c8db.png
[Logistic Regression]: https://blog.csdn.net/qq_40587575/article/details/80877008
[Decision Tree]: https://blog.csdn.net/qq_40587575/article/details/80889223
[KNN_K]: https://blog.csdn.net/qq_40587575/article/details/80900677
[SVM]: https://blog.csdn.net/qq_40587575/article/details/80955436
[Link 1]: https://blog.csdn.net/qq_40587575/article/details/81007236
[Omega]: https://private.codecogs.com/gif.latex?%5COmega
[B_1_ B_2_cdots B_n_Omega]: https://private.codecogs.com/gif.latex?B_1%2C%20B_2%5Ccdots%20B_n%25u4E3A%5COmega
[A_A_Omega _A_cap _B_1_cup B_2_cup_cdots B_n_ _AB_1 _cup AB_2 _cup _cdots AB_n_ P_A_P_B_1_P_A_B_1_cdotsP_B_n_P_A_B_n]: https://private.codecogs.com/gif.latex?A%3DA%5COmega%20%3DA%5Ccap%20%28B_1%5Ccup%20B_2%5Ccup%5Ccdots%20B_n%29%5C%5C%20%3DAB_1%20%5Ccup%20AB_2%20%5Ccup%20%5Ccdots%20AB_n%5C%5C%20P%28A%29%3DP%28B_1%29P%28A/B_1%29%5CcdotsP%28B_n%29P%28A/B_n%29
[8ed5a02ba5e9856a561260f2_pn_8_o_jpg_6_md5sum_70f9cbdaa8e739f584a4e83b46b54fd3_sign_d885d382d7_png_44692-54685_jpg_524661-612874]: /images/20220521/5d94c173945b47f882ebd8657bf7ca0d.png
[8ed5a02ba5e9856a561260f2_pn_11_o_jpg_6_md5sum_70f9cbdaa8e739f584a4e83b46b54fd3_sign_d885d382d7_png_63267-65522_jpg_818745-898101]: /images/20220521/4ac4ec24655e4c349bb35e227058f59b.png
[8ed5a02ba5e9856a561260f2_pn_12_o_jpg_6_md5sum_70f9cbdaa8e739f584a4e83b46b54fd3_sign_d885d382d7_png_65523-65765_jpg_898102-978067]: /images/20220521/0817bb936ba14176bc0776dd11438a87.png
[Link 2]: https://wenku.baidu.com/view/7beebb3731b765ce04081448.html?from=search