Mini-Course in Eye Movement Analysis With Hidden Markov Models (微課程：眼動實驗經濟學專題)

NTU (Fall 2019)

Time: 2019/8/12-16, 2-5pm, at Social Sciences 609 (社科研609)

Office Hour: 5-5:30pm after class or by Email appointment

Course Syllabus

Co-Instructor: Antoni Chan (陳萬師, abchan "at" cityu.edu.hk)

Co-Instructor: Janet Hsiao (蕭惠文, jhsiao "at" hku.edu.hk)

Course Registration Survey

Software 1: Matlab EMHMM Toolbox (to run EMHMM)

Sortware 2: Pre-Processing (from raw data to eye fixations)

Structure and Activities: (Please bring your own laptops and install Matlab and Statistical Toolbox before coming to class!)

1. [8/12]: Introduction to EMHMM and EMHMM Tutorial (my notes)

In the first half of the class, we will introduce current methods in eye movement data analysis to illustrate the advantages of the EMHMM method using face recognition research as an example. We will also briefly introduce EMHMM with co-clustering and Eye Movement analysis with Switching Hidden Markov Models (EMSHMM) so that students can choose to use them for their projects.

In the second half of the class, we will provide an EMHMM Matlab Toolbox tutorial with sample data for students to practice using the toolbox on their own laptops.

2. [8/13]: Using EMHMM in Experimental Research (my notes)

In the first half of the class, we will present an EMHMM simulation study and provide recommendations for using EMHMM in experimental research. In the second half of the class, students will be provided with a sample experiment file in Eyelink Experimental Builder and learn to develop their own mini experiment.

The alternative experimental software is Matlab PTB-3, and here is a tutorial.

3. [8/14]: EMHMM with Co-Clustering and Hands-On Data Collection

In the first half of the class, we will introduce EMHMM with co-clustering using a scene perception task as an example with a short tutorial.

In the second half of the class, students will collect data for their mini-experiment for the project presentation on the last day.

4. [8/15]: EMSHMM and Hands-On Data Analysis

In the first half of the class, we will introduce Eye Movement analysis with Switching Hidden Markov Models (EMSHMM) using a decision-making task as an example with a short tutorial.

In the second half of the class, students will perform data analysis for their mini-experiment and prepare for the project presentation on the last day.

5. [8/16]: Project Presentation