BUS459 Predictive Analytics with R
- Linear regression and classification
- Resampling methods
- Model selection
- Non-linear models
- Tree-based methods
- Unsupervised learning
This course is an introduction to statistical learning and its application to help making economic decisions. Statistical learning is the science of extracting important patterns from data, patterns that can inform a decision maker. Particularly, the course focuses on predictive analytics, the activity of predicting future events or predicting an unobserved property of an individual, a company or some other unit. The course will contain applications on business data. The students will work on those using the statistical software R. Examples of applications are statistical fraud detection and market basket analysis.
After completing the course the students should know the central ideas of statistical learning and predictive analytics and have:
the practical skills to
- implement the most common techniques to real-world prediction problems in R
- evaluate predictions and deal with the problem of overfitting
- appropriately choose between a set of statistical learning methods for a given problem
and the knowledge to
- formulate a research problem for a master thesis
- structure a master thesis
- understand the ethical issues in collecting and interpreting research data
Lectures combined with computer labs.
Basic skills in statistical inference. Basic skills in R.
The final grade will be based on two written group project reports, resulting in one grade. The members of the group must therefore be the same on both projects. The groups should consist of 2, 3 or 4 students.
James, Witten, Hastie and Tibshirani (2013) "An introduction to statistical learning with applications in R".
- ECTS Credits
- Teaching language
Professor Jonas Andersson, Department of Business and Management Science