At the end of this course, you can:
- Apply elementary techniques from orthodox and Bayesian statistics
- Evaluate empirical experimental methods and results
- Implement a (cognitive) theory in a computer simulation (cognitive model) and relate the model results to empirical results
- Report in writing on methods and results from empirical studies and related computer simulations.
The level that you will achieve is that you can make informed decisions about what techniques and methods you want to apply in future projects, such as a Bachelor’s project. Of course, there are many more methods and techniques that you might want to apply beyond what we can cover during the course. We therefore focus on some of the basic principles, so you can apply and extend these more easily at a later time relatively independently. Throughout the course we will emphasize methods and techniques that are relevant for research and applications from an AI perspective.
At a more general level, this course is one of the final courses of your Bachelor's degree in Artificial Intelligence. Therefore, we will also help you to prepare for conducting your own Bachelor research project. For example, we will encourage you to look up and read preceding Bachelor theses, to investigate what research topics are available, and to brainstorm about your own research topic.
People make decisions based on data. This holds for individuals, industry, governments, and scientists. In everyday life you might for example decide whether to go to class by bike or by bus, depending on your knowledge of travel time: how frequently are busses running and is a bus ride shorter than a bike ride? Industry uses data to gain insights in the interests of their customers, or how strong competitors are. Government calculates and tests what the impact of new policies is on society to decide on an effective policy. And finally, scientists use data to develop theories about the many facets of life and the universe.|
To later function properly in your everyday life and in your professional career, you should know how to derive valid and reliable conclusions based on data. To do this, you need to know among others how to collect reliable data (using experimental methods), how to analyze this data (using statistics), and how to communicate these results effectively. These results should be placed in the context of theory (and more generally: in the context of knowledge). Characteristic for AI students and professionals is that they often use computer models and computer simulations to approximate reality. However, how do you then compare a result from a simulation with data from real people? In this course you will learn this and more.
Examination & relationship between grading and course objectives
Examination consists of 4 components:
In the written exam and the (homework) assignments, you will practice statistical techniques and reporting of statistical techniques (course objectives 1 and 4). In the computer lab sessions and final report, you also practice with experimental methods, modeling, and reporting these (course objectives 2, 3, and 4).
- All computer lab assignments are signed off in time (not graded individually)
- A written exam in which theoretical knowledge is tested. The weight of this test is 50% of your final grade. The minimum grade is a 5.5.
- There are periodical homework-assignments that you can hand in related to the workgroups (“werkcolleges”).
- A report in the form of a short scientific article. In this report you cover among others the method of your experiment, analysis of your data, explain the structure of your model, analysis of model data, and you discuss the implications and limitations of your results. The weight of this report is 50% of your final grade. The minimum grade is a 5.5.
The course has three different forms of teaching.
Computer labs Students work on a research project throughout the course. They analyze data of a psychological experiment, develop a cognitive model and test how well the cognitive model describes and explains the empirical data. The results will be written up in a final report. During the labs, students apply the knowledge that they gained during the lectures.
Workgroups (“Werkcolleges”) During the workgroups, students practice with statistical concepts from the lectures. They later also apply these concepts during the computer labs and during the final exam.
Plenary lectures During the plenary lectures we will cover the necessary theoretical background to topics that you will apply and practice during the computer labs and workgroups.
- The lectures by Chris Janssen will mostly focus on methods, including explanation of the research topic (multitasking) and the development and analysis of cognitive models – with an emphasis on functional, process-oriented models. These lectures will be in Dutch, unless non-Dutch lecturers, student assistants, or students attend.
- The lectures of Benjamin Rin will mostly focus on different types of statistics. This will include the difference between Bayesian and more orthodox statistics. These lectures will be in English.
|Some experience with programming is highly desirable. We assume that you are familiar with basic programming terminology such as variable, function, if/else, while and for. In the course we work with the software package called R. Knowledge of and experience with R is not needed before you start the course. The basics of programming in R are covered in the course. |
If you are a student from an external program, then it is also desirable that you are interested in modeling and other AI topics.
|Software R: http://www.r-project.org/|
Aspecten van academische vorming
| • Academisch denken, werken en handelen • Schrijven (algemeen) - diverse typen teksten plannen, schrijven, herschrijven en afwerken • Kennis hanteren in een bredere context • Hanteren van wetenschappelijk instrumentarium |