MØLBA3004 Business Forecasting
- Number of credits7,5
- Teaching semester2024 Spring
- Language of instructionEnglish
- CampusLillehammer
- Required prerequisite knowledge
Recommended prerequisites: MØLBA3001 Data Engineering (datahåndtering og analysedesign) and LDBA200 Applied Programming (anvendt programmering).
The topics covered are:
- Time series data patterns and decomposition
- Moving averages and exponential smoothing methods
- Dynamic regression models
- ARIMA models
- Multivariate time series models (e.g., vector autoregression)
- Advanced forecasting methods (e.g., Neural network and State Space models)
- Forecast combinations
- Loss functions, forecast accuracy and forecast optimality
Learning Outcome
Upon completion of the course, the candidate shall:
- Have advanced knowledge of time-series data patterns and how to decompose them into trend, cycles, season, and random noise (k1)
- Have specialized knowledge of advanced time series models commonly used in business forecasting (k2)
- Explain and exemplify how forecasts generated from various methods can be compared and ranked based on a range of accuracy measures (k3)
- Demonstrate limitations of time series models and (lack of) forecast robustness under unstable conditions (k4)
- Have in-depth knowledge of the key findings from recent research on forecast accuracy of different models applied to business or economic data (k5)
Upon completion of the course, the candidate shall be able to:
- Decompose time series data into structural components using appropriate techniques (f1)
- Generate time series forecasts using the techniques covered in the course (f2)
- Combine individual forecasts to improve accuracy (f3)
- Evaluate forecast accuracy using a range of accuracy measures (f4)
Upon completion of the course, the candidate shall be able to:
- Plan and manage forecasting projects which involve the topics covered in the course (g1)
- Recommend forecasting techniques that are suited in a range of business applications (g2)
- Recognize limitations of forecasting models and anticipate conditions under which they break (g3)
The following teaching methods are used:
- Lectures
- Problem solving sessions
- Tutorial videos
- Case studies
- Quizzes
- Mandatory homework assignments must be handed in before each teaching module (a total of 4). These will be combinations of practical and theoretical exercises covering key topics in the course.
- Three out of four homework assignments must be passed to be allowed to take the exam.
- Attendance on at least 50% of the courses lectured teaching.
Four-hour individual digital exam under attendance.
Graded A-F, where E is minimum for passing the exam.
| Form of assessment | Grading scale | Grouping | Duration of assessment | Support materials | Proportion | Comment |
|---|---|---|---|---|---|---|
Written examination with invigilation | ECTS - A-F | Individual | 4 Hour(s) |
| 100% |
Reading list
No reading list available for this course