To investigate the effect of affective feedback on the safety-critical automated system users’ emotional behaviors and regulation, developed a game simulating a UAV navigation task where users received feedback with emotional content. The project demonstrated that providing affective feedback decreases negative emotions, and increases trust.

Technology used. Unity 3D, C#, Python, iMotions

Theoretically compared four famous parallel hierarchical clustering algorithms and discussed their strengths and weaknesses; and made use case recommendation. The algorithms are Olson 1995; Du et. al. 2005; Rajasekaran 2005; and Li et. al. 2007. The project was supervised by Prof. Sanguthevar Rajasekaran.

Applying a simple logistic regression model on the publicly available UCI human activity recognition (HAR) dataset, I predicted six basic activities sitting, standing, lying, walking, walking downstairs, and walking upstairs. I used only 12 features generated from accelerometer and gyroscope signals which achieved over 75% accuracy.

Technology used. R

Code. GitHub link.