BED4 Decision Modelling and Analysis
This course provides an introduction to how business decision problems can be analyzed using mathematical models and computer tools. Some examples of decision problems that will be covered are:
- How to find an optimal product mix when resources are scarce?'
- How to optimally select suppliers in a tender auction?
- How to construct an optimal work schedule when demand varies over time?
- How to create an optimal investment plan when supply of capital is limited?
- How to determine an optimal portfolio of stocks with different levels of return and risk?
- How to forecast demand based on historical data?
- How to utilize a marketing budget efficiently?
- How to determine an optimal inventory policy, i.e., how often and how much should we refill the inventory?
- How much should we order when demand for a product is uncertain?
- How to determine the sequence and allocation of tasks along a production line with multiple work stations?
- How to determine an optimal transportation plan for a supply chain?
- How to determine the optimal location of production and storage facilities in a supply chain?
The choice of models in a particular situation depends on the properties of the decision problem that we are analyzing. All the models that will be studied can be analyzed using computer tools, and in this course we will be using Analytic Solver, which comes as an add-in in Excel. We will start with linear models (linear programming), which can be solved and analyzed relatively easily. Later in the course we will study problems with either/or-decisions, which requires models with integer variables, as well as non-linear decision models. Throughout the course we will emphasize interpretation of analysis results, as well as practical implications for businesses.
An important topic in our course is handling of uncertainty in decision situations. We will look at how we can make good forecasts, as well as how simulation models can be used to evaluate consequences of different decision alternatives under uncertainty. We will also study how risk attitudes can affect the choice between decision alternatives, e.g., how risk averse decision making can be modeled. Finally, we will look at how the value of additional information can be computed when future outcomes are uncertain, and how decision trees can be used to structure complex decision situations.
The students know how to analyze different business decision problems using mathematical models and computer tools.
- can formulate decision problems mathematically and implement the models using computer tools.
- can graphically solve and do sensitivity analysis for an LP problem with two decision variables.
- can use results from sensitivity reports for LP problems, including shadow prices and slack values, in economic analyses.
- can use integer variables to model various discrete choices, such as either/or decisions.
- know computational problems related to the analysis of integer models and non-linear models.
- can determine when it is necessary to use integer models or non-linear models.
- know and are able to use basic models for forecasting, including models that take into account trend and seasonal variations.
- can use simulation models to quantify the consequences of decisions under uncertainty.
- are able to calculate and understand confidence intervals for estimates from simulation models.
- know the most central decision criteria for analyzing decisions under uncertainty, including models that take into account different types of risk attitudes.
- can use decision trees to analyze simple decision problems with expected value as decision criterion.
- understand and can calculate the value of perfect/imperfect information based on expected value as decision criterion.
The students can communicate with experts on modeling and analysis of business decision problems.
Credit reduction due to overlap
4 hour written school exam.
A - F
- ECTS Credits
Autumn. Will be offered Autumn 2023.
Professor Endre Bjørndal, Department of Business and Management Science
Professor Mette Bjørndal, Department of Business and Management Science