Course talk:CPSC522/Improving Human Behavior Prediction in Simultaneous-Move Games

From UBC Wiki

Contents

Thread titleRepliesLast modified
Suggestions117:10, 23 April 2016
Questions regarding your page116:57, 23 April 2016
Comments and critique116:52, 23 April 2016
Feedback on Improving Human Behavior Prediction in Simultaneous-Move Games116:50, 23 April 2016

Suggestions

Hi Adnan,

I enjoy reading your page.

Here are a few suggestions:


>Related Literature

It is better to cite the related literatures.

>It might be better not to put texts and formulas in figures. If you explain texts of the figure by yourself, even without the formulas, it will be more understandable for people of different research areas.

You may want to add figure 2 after starting to explain about that.

I know that cross validation works better than splitting the data-set to train and test sets. But I think cross validation is used for selecting parameters and after specifying the parameters, the model with the defined parameters will be built based on the large amount of data. Why have you used cross validation and have you tried to build the model based on the whole data (or large amount of data)?

BahareFatemi (talk)20:08, 20 April 2016

Thank you Bahare for pointing out these issues. I have gone through the page and added citations where it was missing. I have also included a brief explanation of figure 2 before adding it. One of primary authors of the working paper on which this page is based recommended using cross validation since it would generalize better to new data.

AdnanReza (talk)17:10, 23 April 2016
 

Questions regarding your page

Hi Adnan,

Awesome work here and I appreciate your effort into this page.

Just one question, why are the evaluation graph picture all so vague? Can you help to provide the pesudo code or source code so that I can better understand the process. Thanks

Regards Arthur

BaoSun (talk)08:05, 19 April 2016

Thanks Arthur for pointing that out. I have included a link to the python code for the nelder mead simplex algorithm used here. I have also tried to better explain some of the graphs showing the results.

AdnanReza (talk)16:57, 23 April 2016
 

Comments and critique

Hi Adnan, Firstly, let me thank for your project and your contribution in CPSC 522 Wiki pages. I really like your page and find your topic of project interesting and you have done a fantastic job! I like to share my opinion and give you my comments, maybe they might help you.

1) The only super minor thing I would say is that the positioning of some of the figures are weird and makes look of some of the subsections unusual, especially the figures that are on the left side of the screen. But other than that, I loved it!

Thanks for your page it was a pleasure to read it and I hope these comments can help you improve your contribution. Mehrdad Ghomi

MehrdadGhomi (talk)03:06, 21 April 2016

Thanks Mehran. I will change the alignment of the figures.

AdnanReza (talk)16:52, 23 April 2016
 

Feedback on Improving Human Behavior Prediction in Simultaneous-Move Games

Hi Adnan, Nice work on the page, specially the usage of examples. I have the following things to mention:

  • Some of the paragraphs are a bit long. Breaking down into smaller ones would be helpful
  • Some of the figures and/or equations can be explained a little bit more for the naive reader.

Finally, I really appreciate your hard work and contribution in development of the WIKI.

MDAbedRahman (talk)07:51, 21 April 2016

Thank you Abed. I have tried to simplify some of the paragraphs in order to make it more readable.

AdnanReza (talk)16:50, 23 April 2016