Operational
Analytics- An Over
View
BY
PRUDHA PALLAVI PENTAPATI
Descriptive Analytics
 What is a fundamental Operational Problem?
 Time line of events
Steps before the Solution
 Analyse the past data
 How much should you order
 News Vendor Problem
 Time’s Magazine Aprroach
 Broader application to the problem
Forecasting
 What is Forecasting?
 Characteristics of Forecasts
 Probablity Distributions in Forecasting
 Model for future Demand
Making Best decisions with low
uncertainity
 A resource allocation Example
 Converting a verbal problem application into an algebraic model
 Algebraic model to spreadsheet implementation
 Matching demand and supply across space
Decision Variables
 What are decision variables?
 Optimizing Production
 Build an Objective Function
 Working around constraints
 How to Identify Key Performance Indicators
Risk and Evaluation of Alternatives
 Making Decisions in Low-Uncertainty vs. High-Uncertainty Settings
 Connecting Random Inputs and Random Outputs
 Modeling Random Variables using Scenarios
Decision Trees
 What is a decision tree
 Structure of a decision tree
 Understanding a decision tree
Approach to Evaluate Options using
Decision Trees
 “Maxi-min” strategy
 Choose the actions that maximize the minimum outcome
 Avoids bad outcomes…and…ignores the possibility of good outcomes
 “Risk averse” strategy
 “Maxi-max” strategy
 Choose the actions that maximize the maximum outcome
 Seeks good outcomes...and…ignores the possibility of bad outcomes
 “Risk seeking” strategy
 Maximize the expected value of the outcomes
 Gives equal weight to good and bad outcomes
 “Risk neutral” strategy
Decision Trees in Practice
 IDEA is a small example designed to convey the essential ideas
 In practice, decision trees are used to evaluate a wide range of complex
problems. You can find examples in articles published in Interfaces m R&D
licensing
 Credit Scoring ,Polio Eradication
 There exists software to help manage and analyze large decision trees m
core-feature, single user, products: e.g., Tree Plan (free trial) massive-
feature, enterprise-use products: Decision Tools, Logical Decisions

Operational analytics overview

  • 1.
  • 2.
    Descriptive Analytics  Whatis a fundamental Operational Problem?  Time line of events
  • 3.
    Steps before theSolution  Analyse the past data  How much should you order  News Vendor Problem  Time’s Magazine Aprroach  Broader application to the problem
  • 4.
    Forecasting  What isForecasting?  Characteristics of Forecasts  Probablity Distributions in Forecasting  Model for future Demand
  • 5.
    Making Best decisionswith low uncertainity  A resource allocation Example  Converting a verbal problem application into an algebraic model  Algebraic model to spreadsheet implementation  Matching demand and supply across space
  • 6.
    Decision Variables  Whatare decision variables?  Optimizing Production  Build an Objective Function  Working around constraints  How to Identify Key Performance Indicators
  • 7.
    Risk and Evaluationof Alternatives  Making Decisions in Low-Uncertainty vs. High-Uncertainty Settings  Connecting Random Inputs and Random Outputs  Modeling Random Variables using Scenarios
  • 8.
    Decision Trees  Whatis a decision tree  Structure of a decision tree  Understanding a decision tree
  • 9.
    Approach to EvaluateOptions using Decision Trees  “Maxi-min” strategy  Choose the actions that maximize the minimum outcome  Avoids bad outcomes…and…ignores the possibility of good outcomes  “Risk averse” strategy  “Maxi-max” strategy  Choose the actions that maximize the maximum outcome  Seeks good outcomes...and…ignores the possibility of bad outcomes  “Risk seeking” strategy  Maximize the expected value of the outcomes  Gives equal weight to good and bad outcomes  “Risk neutral” strategy
  • 10.
    Decision Trees inPractice  IDEA is a small example designed to convey the essential ideas  In practice, decision trees are used to evaluate a wide range of complex problems. You can find examples in articles published in Interfaces m R&D licensing  Credit Scoring ,Polio Eradication  There exists software to help manage and analyze large decision trees m core-feature, single user, products: e.g., Tree Plan (free trial) massive- feature, enterprise-use products: Decision Tools, Logical Decisions