Week 10 Readings and Homework

Due Nov 17 10am:

(1) Prepare slides for your "Teach the class something about work related to your project" presentation for Nov 18 and upload them by Nov 17 10am.  Each student will teach for 7 minutes on something you've learned about while reading related work for your project.  Ideally it is something related to the affective computing elements of your project -- it can be highlights from a background paper you read, theories related to affective elicitation or measurement, anything related to your project and affective computing. You will have 7 minutes to present and 2 mins for Q&A.  Please be *very tutorial* - clarity will contribute to your grade for this exercise.  Do not assume that your fellow students have read what you've been reading, and note that some do not have machine learning backgrounds, so you may need to explain some basics.

(2) Remember the milestone of spending ~12 hours week total on this class (~10 outside of class when we meet).  Since we do NOT meet for class Nov 11 (Veteran's Day Holiday) you have two weeks before this (last) homework is due.  Jot a few sentences about how you spent your time these two weeks for advancing your project.  You can work straight on moving your project/study/data analysis forward, or if you can't start your study because it's held up by COUHES, thengo ahead and start drafting your project report and final presentation slides, putting in all of the content that you already know.

Upload a draft report with the following based on what the final report needs:

(a) Choose a template.  You can choose a conference you might want to attend (if it is submitted/accepted) and use their template.  If you do this and their template is short (e.g. 4 pages) then you can feel free to add supplemental pages showing more of your data/findings if you wish.  Some previously used conference templates include CHI, ICMI, ACII, ICASSP, etc.   If you don't know, that's also fine, and you can just use whatever format fits your style.  A typical non-conference final report would be about 6-12 pages single column ( more pages go with more data/graphs to show).

(b) Collect all of the readings you've already put into earlier homeworks for this class, and put them into the bibliography.  Also write a background section and put in a few sentences to describe each related work listed in the bibliography.

(c) Write up your experiment design or basic Methods section of your project.

(d) Leave these sections empty for now:  Abstract, Introduction, Results, Discussion/Conclusions/Future Work.  You will do these later (although if you have time to do the Introduction now, feel free.)

Your final in-class presentation will also need slides that frame your project goals and describe your methodology.   Please do as much as you can these two weeks and report here how your time was spent.

(3) Try this fun demo (sorry, due to a recent change at Google, it presently works only on an iphone or iOS device): While you hold still, and try to relax, turn the browser in your smartphone to the demo at try.globalvitals.com Links to an external site.. Follow it's instruction and see if you can read your heart rate using the phone held (a) against your chest and (b) just resting the phone on the top of your thigh near your hip while sitting still relaxing. This demo, made by MIT Media Lab alum Javier Hernandez, shows your phone (even in your pocket) can read your heart rate when you're "still" by processing rhythms in very subtle motions detected by the phone.  Did it work for you in either or both positions?  While heart rate can also change with non-emotional activity,  it is an easily measured physiological response that also accompanies many emotional state shifts.