当前位置:网站首页>2022 8th "certification Cup" China University risk management and control ability challenge
2022 8th "certification Cup" China University risk management and control ability challenge
2022-07-07 12:00:00 【Enthusiastic netizen Mr. Minamata】
The start time of this risk control cup is the beginning of summer , This game can also be used as a practice match of the National Games for running in between teammates . The theme of this risk control competition is the current car loan , The title gives the data set of the international market . The title is as follows ,
According to the title requirements , We can see that the topic generally requires us to carry out two steps , One 、 Screening of relevant indicators , Two 、 Establishment of prediction model . Here we explain these two steps separately .
The most critical step before the index processing is our processing of the given data , Because the given data may have outliers , We need to deal with the vacancy value accordingly , The relevant processing code is also summarized for you .
One 、 Relevant index screening
According to the needs of the topic , The first step is to screen out relevant indicators , In order to facilitate our next step of model establishment, solution and related analysis . For the selection of indicators , Here I think there are two ways to screen , Both objective and subjective . namely , We can artificially screen some indicators as preliminary indicators ; At the same time, we can also use correlation analysis 、 Causal analysis screens key indicators . I will share some related codes for you later , I hope I can help you .
Two 、 Establishment of prediction model
Predict whether the lender will default , This problem itself is a problem of establishing predictions based on data . I think there can be multiple linear regression prediction , Time series prediction , Grey prediction , Fuzzy prediction, etc . There is a better way to establish here is to choose two or three prediction models , Finally, compare the accuracy of the results . In this way, our model will be very complete , At the same time, effectively increase the length of the article . The code of the prediction model will be supplemented later .
Final , I wish you all good results in this competition , Be able to run in with your teammates more , Create more brilliance in the National Games .
边栏推荐
- 【神经网络】卷积神经网络CNN【含Matlab源码 1932期】
- 【纹理特征提取】基于matlab局部二值模式LBP图像纹理特征提取【含Matlab源码 1931期】
- About how to install mysql8.0 on the cloud server (Tencent cloud here) and enable local remote connection
- Excel公式知多少?
- How to connect 5V serial port to 3.3V MCU serial port?
- SwiftUI 4 新功能之掌握 WeatherKit 和 Swift Charts
- Le Cluster kubernets en cours d'exécution veut ajuster l'adresse du segment réseau du pod
- 2022 年第八届“认证杯”中国高校风险管理与控制能力挑战赛
- Complete collection of common error handling in MySQL installation
- Talk about SOC startup (VI) uboot startup process II
猜你喜欢
Cmu15445 (fall 2019) project 2 - hash table details
Camera calibration (1): basic principles of monocular camera calibration and Zhang Zhengyou calibration
[extraction des caractéristiques de texture] extraction des caractéristiques de texture de l'image LBP basée sur le mode binaire local de Matlab [y compris le code source de Matlab 1931]
Superscalar processor design yaoyongbin Chapter 9 instruction execution excerpt
从工具升级为解决方案,有赞的新站位指向新价值
Superscalar processor design yaoyongbin Chapter 8 instruction emission excerpt
《通信软件开发与应用》课程结业报告
Common locking table processing methods in Oracle
[filter tracking] strapdown inertial navigation pure inertial navigation solution matlab implementation
[shortest circuit] acwing 1127 Sweet butter (heap optimized dijsktra or SPFA)
随机推荐
从工具升级为解决方案,有赞的新站位指向新价值
HCIA复习整理
问下flinkcdc2.2.0的版本,支持并发,这个并发是指多并行度吗,现在发现,mysqlcdc全
竟然有一半的人不知道 for 与 foreach 的区别???
What is high cohesion and low coupling?
R语言使用quantile函数计算评分值的分位数(20%、40%、60%、80%)、使用逻辑操作符将对应的分位区间(quantile)编码为分类值生成新的字段、strsplit函数将学生的名和姓拆分
Detailed explanation of debezium architecture of debezium synchronization
Present pod information to the container through environment variables
Flet教程之 18 Divider 分隔符组件 基础入门(教程含源码)
powershell cs-UTF-16LE编码上线
. Net Maui performance improvement
R語言使用magick包的image_mosaic函數和image_flatten函數把多張圖片堆疊在一起形成堆疊組合圖像(Stack layers on top of each other)
【纹理特征提取】基于matlab局部二值模式LBP图像纹理特征提取【含Matlab源码 1931期】
MATLAB實現Huffman編碼譯碼含GUI界面
[data clustering] realize data clustering analysis based on multiverse optimization DBSCAN with matlab code
Talk about SOC startup (VI) uboot startup process II
Camera calibration (2): summary of monocular camera calibration
总结了200道经典的机器学习面试题(附参考答案)
[neural network] convolutional neural network CNN [including Matlab source code 1932]
超标量处理器设计 姚永斌 第10章 指令提交 摘录