Introduction to Data Analysis and Decision Making – Chapter 1

 

I.                   Introduction

·        Businesses increasingly have employees utilize quantitative methods to transform data into useable information

·        Objective is to teach how to use quantitative methods to analyze data and make decisions

 

II.                Overview of the Book

Methods

·        Combine statistics and management science

A.            Three general themes

-         Data Analysis

·        Data description

·        Data inference

·        Search for relationships

-         Decision Making

·        Optimization techniques for problems with no uncertainty

·        Decision analysis for problems with uncertainty

·        Structured sensitivity analysis

-         Dealing with Uncertainty

·        Measuring uncertainty

·        Modeling uncertainty explicitly into the analysis

B.            Contents of the Book

-         Ch 2,3 – summarizing information in data sets

-         Ch 4 – probability and probability distributions

-         Ch 5 – normal and binomial distributions

-         Ch 6 – probability in decision maker under uncertainty

-         Ch 7,8,9 – sampling and statistical inference

-         Ch 10 – statistical process control (quality control)

-         Ch 11,12 – regression analysis

-         Ch 13 – times series analysis and forecasting

-         Ch 14,15 – spreadsheet optimization and linear programming

-         Ch 16 – computer simulation models

C.            Software

-         Built in excel features – tutorial.htm

-         Solver add-in – uses algorithms for spreadsheet optimization

-         Stat pro add-in – generates output quickly and easily

-         RandFns add-in – generates random value

-         SolverTable add-in – used for sensitivity analysis

-         Decision Tools suite – decision tools add-in

·        @Risk – run multiple replications of spreadsheet simulation

·        PrecisionTree – analyze decision problems with uncertainty by drawing decision trees

·        TopRank – sensitivity analysis to see which inputs most affect output

·        BestFit – determine most appropriate probability distribution

·        RiskView – drawing tool that complements @Risk

-         Software guide – figure 1.2

 

III.             Sampling of Examples

1.      Scatterplot

2.      Decision Tree

3.      Statistical Inference

4.      Chart comparison

5.      Scatterplots and Multiple Regression

6.      Time Series Plot

7.      Optimization example

8.      Simulation model

 

IV.              Modeling and Models

A.                 Graphical Models

·        Most intuitive and least quantitative type of model

·        Portray graphically how different elements of a problem are related

B.                 Algebraic Models

·        Specify a set of relationships in a very precise way

·        Can be stated concisely and with great generality

·        Require an ability to work with abstract mathematical symbols

C.                 Spreadsheet Models

·        Intuitive given the instant feedback

·        Must be well designed and well documented to be effective

·        Strength is the flexibility of spreadsheet models

D.                 The 7 Step Modeling Process

                                                                                                                                       i.      Define the problem

                                                                                                                                     ii.      Collect and summarize data (data analysis)

                                                                                                                                    iii.      Formulate a model

                                                                                                                                   iv.      Verify the model

                                                                                                                                     v.      Select one or more suitable decisions (decision making)

                                                                                                                                   vi.      Present the results to the organization

                                                                                                                                  vii.      Implement the model and update it through time