Please enable JavaScript.
Coggle requires JavaScript to display documents.
Decision Models - Coggle Diagram
Decision Models
Two types
Descriptive & Exploratory
Prospective
[EX] Problems that are concerned with predictions and estimations
Perspective
[EX] Exploratory data analysis
Prescriptive analysis
One tries to find the best course of action or decision
[EX] Optimization problems
Phase 3: Sensitivity (What-if analysis)
How do changes in the values of the model parameters affect the model solutions?
[EX] When performing a break-even analysis, one would want to know how changes in unit prices or in unit costs would affect the solution of the model
The sensitivity analysis of a model is going to shed light on the question of whether the model solution is unique.
[EX] An appropriate sensitivity analysis of a model should be able to detect the ranges of values of model parameters that give rise to a unique solution and ranges that would cause multiple solutions.
Phase 1: Model design
A model would involve 3 quantities whose values would influence the outcome of that model
Model variables
Uncontrollable variables
They are changing quantities that cannot be controlled by the decision-maker. (In some cases, they may be random values whose behavior would be according to some probability distribution)
[EX] demands, return of investment
Controllable variables
(Decision variables)
Their value can be chosen by the company's decision makers
They are always the main fruit of a model, their values would constitute the model solutions after the model has been solved.
Model objective
The type of objective in a model depends on the type of the model.
[EX] Descriptive model such as break-even analysis, the objective of it is to attain a zero profit
[EX] the objective of a portfolio optimization model is to maximize the return of investment or to minimize the associated risk.
Regardless of the objective, the objective of a model should always be mathematically expressed in terms of the decision variables and the parameters of that model.
Model parameters
Whose values are influenced by the factors that are outside of the domain of the control of the decision-making
[EX] Unit prices, unit costs, distances, wages...etc.
[EX] the state of the economy and labor laws are two such factors that would influence the parameters of a model
Phase 2: Model solution/
Obtaining meaningful solutions
The solutions of the model are either specific values of the decision variables or a range of such values.
3 Questions that the decision maker needs to answer in this phase
If the model has a solution, is that unique? Does the model has multiple solutions?
This may sometimes happen when the same objective can be satisfied by using different combinations of the values of the decision variables.
How is the performance of our solver when we are trying to solve the model?
Does the solution?