Please enable JavaScript.
Coggle requires JavaScript to display documents.
LIL - Business Analytics Foundations: Predictive, Prescriptive, and…
LIL - Business Analytics Foundations: Predictive, Prescriptive, and Experimental Analytics & Descriptive, Exploratory and Explanatory Analytics
Business Analytics
-
-
Steps
Descriptive - What happened?
Exploratory - What is going on?
Explanatory - Why did it happen?
Predictive - What will happen?
Prescriptive - How do I take advantage?
Experimental - How well will it work?
Predictive Analytics
-
-
Machine Learning Types
Unsupervised
Clustering
- k-Means
- k-Nearest Neighbors
-
Supervised
Classification
- Decision trees
- Random forest
- Naïve Bayes
- Neural networks
- Support vector
-
-
Best Practices
Focus on business gain and ROI, not pure model performance
-
-
-
Test, retest and periodically test model accuracy w new sets of data
-
Experimental Analytics
Why?
The goal of experiemental analytics is to test a hypotesis or alternative to understand actual performance on the field
Steps
- hypotesis
- design experiment
- execute experiement
- analyze results
- choose option
- iterate if required
-
-
Prescriptive Analytics
Why?
The goal of prescriptive analytics is to identify ways and means to take advantage of the findings and predictions provided by earlier stages of analytics
How?
Linear Programming
What?
-
Linear relationships between decision variables
Constraints like budget, time, and human resources are modeled
Decision Analysis
What
Procédures, methods, and tools to analyse business decisions
How?
Create alternatives (including a no-action alternative)
Benefits: monetary and non-monetary
Costs: budget, time, resources
Threats: internal, external, competition
Opportunities: new customers and/or markets
Uncertainty: modeling unknown
Simulation
Manual or automated
All environmental variables should be considered
Run fo multiple options and scenarios
Modify input and see impact on output
Descriptive Analytics
What?
Present facts about the business (performance, etc.)
-
-
-
Exploratory Analytics
-
How?
Process
- Insight needed for specific problem
- Explore, segment and profile
- Share results
- Iterative
-
-
-
-
Explanatory Analytics
What?
- Storytelling w data
- Answer questions
- Present to audiance
- Done by analysts/manager
- Prelude to next action
Steps
- start w a fact
- break down to profiles/segments
- focus on insights
- narrow down to possible causes
How?
-
Graphical tools
E.g. Fishbone diagram
-
-