10. Model, Prediction & Decision (IB533 110.1)
Unit Outlines:
■ Logistic Regression
§ method:glm(… , family = binomial)
§ model fitting & prediction
§ interpreting the coefficients
■ Measures of Model Accuracy
§ Accuracy, Sensibility, Specificity
§ Confusion Matrix and Accuracy Metrics
§ ROC and AUC
■ From Prediction to Decision
§ Dist. of Predicted Probabilities
§ The Critical Threshold
§ Confusion Matrix
§ Payoff Matrix
§ Max. Expected Payoff
§ Assumptions & Simulation
Notebooks:
◆ UNIT10A: Logistic Regression (unit10A.html)
◆ UNIT10B:LogisticReg. – The Coefficients (unit10B.html)
◆ UNIT10C LogisticReg. – Model Accuracy (unit10C.html)
◆ UNIT10D Model Prediction vs. Business Decision (unit10D.html)
Material Download:
◆ Unit10 Share Folder 【 10LogisticReg 】
◆ Unit10 Slides ( 10_LogisticReg.pttx )
◆ Unit10 Codes & Data ( 10_LogisticReg.zip )
Unit10a Personal Assignments : Due at 2022/1/2 (Sun) 23:59
Register to edX and sign in to MIT’s The Analytic Edges course with audit option
Finish Unit3-1 Modeling the Expert: An Introduction to Logistic Regression
Do the quiz in a Rmd, knit the HTML and
upload it to 【 i.html/AnalEdge3 】with filename “<student.id>_AnalEdge2.html“
Unit10b Personal Assignments : Due at 2022/1/9 (Sun,) 23:59
§ Complete the Quiz in unit10C & unit10D
§ unload to 【 ~/i.html/unit10/ 】
§ with file names “_unit10<C|D>.HTML” respectively
Video Recording:
◆ Unit10 Video Recording
References :
◆ Portal of Big Data BA Platform
◆ R: Self Learning Roadmap
◆ John Hopkins: R Programming online course
◆ John Hopkins: R Programming eBook
◆ Harvard: Data Science online course
◆ Harvard: Data Science ebook
◆ From Data to Viz
◆ Probability Cheat Sheet (pdf)
◆ Intro. Probability and Statistics Using R (IPSUR)
◆ Statistical Analysis Using R, UCLA (URL)
◆ CLUSTER ANALYSIS IN R: PRACTICAL GUIDE (URL)
◆ Principal Component Methods in R: Practical Guide (URL)