My notes:
- I really like the opening paragraph with the diagram. I thought is was clear and concise. Having the image made it understandable since you have such a visual project.
- Is there a reason that you are only doing 2D?
- Do you have an example of what a naive way is?
- I always like when data sets are given as a table with the name, description, type, etc for each feature in the set
- Why are you using other people's code and who are they?
- Will you have a validation set?
- Will you do cross-fold validation?
- I think I am left asking what exactly is your end goal on the data. Is it to find what features men and/or women find the most important when picking a partner to date?
- Other questions that I have is, are there any assumptions, like do we assume that everyone is heterosexual, homosexual, bi?
- What do you mean the left 50? physically to the left or do you have a 500 in your entire data set, 450 for training and the ones remaining are for testing?
Overall I thought your paper sounded really good. It was very understandable.