Predictive Analytics - vvvvvvvvvvvvvvv bbbbbbbbbbbbbb cccccccccc PDF

Preview

Summary

vvvvvvvvvvvvvvv bbbbbbbbbbbbbb cccccccccc...

Description

INDIAN INSTITUTE OF MANAGEMENT KOZHIKODE Executive Post Graduate Programme in Management Course Outline Course Code and Course Title Course type Pre-requisites (if any) Course Credit

Predictive Analytics Elective Quantitative Techniques 2

Introduction The recent rise of analytics has resulted in a great demand for business analysts and the trend will only continue to rise. Predictive analytics is one of the important aspects of business analytics. It is all about offering actionable business predictions through interesting and meaningful patterns in data. It has been proven to be incredibly useful in industries such as banking, insurance, telecom, retail, travel, healthcare etc. and has shown immense positive impact on business decision making. Many companies are showing their inclinations towards predictive analytics to thrive and compete against their competitors. The course involves extensive hands-on exercises with different predictive analytic techniques using statistical software R. The course will begin with exploratory data analysis followed by indepth discussion of predictive analytic tools with their applications in various domains. Learning Outcomes/Course Objectives The course has the objective of introducing the participants with the concepts, methods, and techniques of predictive analytics. Participants will also gain the requisite skills in statistical software to perform predictive analytics in real-life business scenarios and use the techniques to interpret model outputs. At the end of the program the participants will be able to  identify business situation where predictive analytics can be applied and the benefits which can be derived  acquire software skill in R and its implementation in predictive analytics  apply predictive analytic tools appropriately  assess and select the appropriate analytic tool to validate the predictive model  analyze the results and communicate the decision to end users Textbooks and Learning Materials Textbook  Seema Acharya (2018). Data Analytics using R. McGraw Hill Education Reference Books  James, G., Witten, D., Hastie, T., &Tibshirani, R. (2013). An Introduction to Statistical Learning: with Applications in R. New York: Springer-Verlag. (web: http://wwwbcf.usc.edu/~gareth/ISL/)



Siegel, E. (2016). Predictive Analytics. Wiley.



Hyndman, R. J. &Athanasopoulos, G. (2016). Forecasting: Principles and Practice. Otexts. (web: https://www.otexts.org/fpp/)

Cases 

Women and Children First on the Titanic(HBSP). Chris A. Higgins and Crystal Ji (2013). (https://hbsp.harvard.edu/product/W13259-PDFENG?Ntt=Women+and+Children+First+on+the+Titanic&itemFindingMethod=Search)



Healthcare Analytics: Vanderbilt University Medical Center- Elective Surgery Schedule (HBSP). Lauren E. Cipriano et al.,(2015 (https://hbsp.harvard.edu/product/W15166PDFENG?Ntt=2.%09Healthcare+Analytics%3A+Vanderbilt+University+Medical+Center+Elective+Surgery+Schedule&itemFindingMethod=Search)



HR Analytics at Scaleneworks (HBSP). Rahul Kumar and U. Dinesh Kumar (2016). (https://hbsp.harvard.edu/product/IMB551-PDFENG?Ntt=3.%09HR+Analytics+at+Scaleneworks++Behavioral+Modeling+to+predict+renege&itemFindingMethod=Search)



Retail Credit Scoring for Auto Finance Ltd. (HBSP). Sujoy Roychowdhury and Srinivas Prakhya (2014). (https://hbsp.harvard.edu/product/IMB467-PDFENG?Ntt=4.%09Retail+Credit+Scoring+for+Auto+Finance+Ltd.&itemFindingMethod= Search)



Demand forecasting for perishable short shelf-life homemade food at iD Fresh Food (HBSP). Raman Narasimhan et al. (2018). (https://hbsp.harvard.edu/product/IMB653-PDFENG?Ntt=5.%09Demand+forecasting+for+perishable+short+shelflife+homemade+food+at+iD+Fresh+Food&itemFindingMethod=Search)

Reading Materials 

How to Design a Business Experiment (HBSP). Oliver Hauser and Michael Luca (2015). (https://hbsp.harvard.edu/product/H02FSL-PDFENG?Ntt=HOW+TO+DESIGN+A+BUSINESS+EXPERIMENT&itemFindingMethod=Searc h)



When to Act On a Correlation and When Not To. (HBSP). David Ritter (2014). (https://hbsp.harvard.edu/product/H00Q1X-PDFENG?Ntt=When+to+Act+On+a+Correlation+and+When+Not+To&itemFindingMetho d=Search)



Can Machine Learning Solve your Business Problem? (HBSP). AnastassiaFedyk (2016). (https://hbsp.harvard.edu/product/H03A8R-PDFENG?Ntt=Can+Machine+Learning+Solve+your+Business+Problem%3F&itemFindingM ethod=Search)



Analytics 3.0 (HBSP). Thomas H. Davenport (2013). (https://hbsp.harvard.edu/product/R1312C-PDFENG?Ntt=Analytics+3.0&itemFindingMethod=Search)



3 Common Mistakes That Can Derail Your Team's Predictive Analytics Efforts (HBSP). Eric Siegel (2018). (https://hbsp.harvard.edu/product/H04KHM-PDFENG?Ntt=3+common+mistakes&itemFindingMethod=Search)



Minding the Analytics Gap (HBSP). Sam Ransbotham, David Kiron and Pamela Kirk Prentice (2015) (https://hbsp.harvard.edu/product/SMR522-PDFENG?itemFindingMethod=Other)

Technology and Software 

R

Other Resources (Journals, Internet Websites) Journals  Computational Statistics and Data Analysis (web: https://www.journals.elsevier.com/computational-statistics-and-dataanalysis/)  The R Journal (web: https://journal.r-project.org/)  Journal of Statistical Software (web: http://www.jstatsoft.org/index) Internet Websites  https://www.informs.org/Community/Analytics  http://analytics-magazine.org/  https://www.kaggle.com/  http://www.r-bloggers.com/  http://blog.revolutionanalytics.com/r/  

http://chance.amstat.org/ http://www.statslife.org.uk/significance

Pedagogy Used/Learning Process The course is a mix of lectures, case discussions, hands on training with statistical software R. The predictive analytic methods/techniques will be discussed in each session led by the instructor with elaborate data analysis followed by student led case discussions in a variety of industry settings. Students must bring their laptop in each class. Homework and assignments will be assigned to the students at regular intervals.

Evaluation Components/Assessment of Student Learning Evaluation /Assessment Tool

Component Percentage

End term Exam

50%

Class Participation

30%

Project

20%

Description These components will evaluate students’ ability to understand various predictive analytics methods and to apply them in the context of business problems. The ability to interpret the R output will also be evaluated.

“Projects” will be assigned to the students in groups. The groups will be decided on the first session itself. Each group will identify a relevant project topic and provide a detailed analysis using the predictive analytics methods discussed in the class. Project will help students demonstrate their ability in handling the complex business problems.

Session Plan Session* 1

Module* Introduction to Predictive Analytics

Topic** Chapter No. Analytics Paradigm, Types of Problems Supplemental Materials in Analytics, Prediction Effect

2

Introduction to R

3-4

Data Visualization and Basic Statistics

5-9

Linear Models: Predictive Analytics using Linear Regression

Installing R, Installing R package, Chapters 1-3 Loading R package, Reading Data Sets in R, Basic Mathematical Operations in R Data Visualization, Descriptive Chapter 4 Statistics, Demo using R Case: Women and Children First on the Titanic Linear regression, Model building Chapter 5 through Dummy, Polynomial and Interaction variables, Model validation, Model Diagnostics, Multicollinearity,

10

11-14

Cross-Validation

Classification Problem: Generalized Linear Models

Model deployment, Demo using R Case: Healthcare Analytics: Vanderbilt University Medical Center- Elective Surgery Schedule Training Error Rate and Test Error Rate, Supplemental Hold-out Sample, Validation Set Materials. Approach, Demo using R Case: Healthcare Analytics: Vanderbilt University Medical Center- Elective Surgery Schedule Logistic Regression, Sensitivity and Chapter 6 Specificity, ROC curve, Demo using R Case: HR Analytics at Scaleneworks

15-16

Time Forecasting

Series

Time Series Components, Modeling Time Series Data using Box– Jenkins method, Forecasting, Demo using R

Chapter 8

Case: Demand forecasting for perishable short shelf-life homemade food at iD Fresh Food End Term Examination

* Tentative ** Each topic will be first demonstrated using caselet. ** Supplemental Materials will be provided, whenever required. Group project: A project proposal containing the project title, objectives, data source, etc., should be submitted for approval on or before the completion of eighth session. Once the topic for the project is decided, each group should start working on the project. Each group should regularly share the progress in the project work with the instructor and submit an interim report at the end of the thirteenth class. The deadline for submission of the final project report will be intimated at the end of the penultimate session. The length of the final project report should be at least four pages but no more than ten pages (A4 size, font size 12). The final report must contain the following:  A title of the project with introduction and motivation for the problem  Data Source(s)  Exploratory Analysis and Descriptive Statistics  Detailed Discussion on the Application of businessanalytics methods  Results  Conclusion Data Source (s), Tables and graphs and R codes should be provided in the “Appendix” section of the report. Each group should submit a soft copy (preferably in PDF format) of the

project report along with the relevant data set through Moodle. Each report must include the names of all the group members along with the group number. Each member of the group should contribute in the preparation of the final report....