search

huizezhang-sherry / paper-ferrn

Other
0 stars 0 forks source link

Several points from the panel #14

Open huizezhang-sherry opened 3 years ago

huizezhang-sherry commented 3 years ago
  1. Gael wrote the following on the first page of the paper: "Explain technical term - or not use"

I suspect this relates to the third paragraph in the introduction section as multiple concepts regarding projection pursuit guided tour are introduced a bit too abrupt

huizezhang-sherry commented 3 years ago
  1. Dan feels that some paragraphs, in the beginning, are correct but not necessary

I suspect this relates to the second paragraph in introduction and the optimisation methods section as they are too broad. After my presentation and more readings on the projection pursuit papers in the 80s, I feel we should go directly into the type of optimisation of our concern - a constrained optimisation with the constraint imposing a geometric/ topological structure on the search space.

Hence I propose to reduce the optimisation methods section with a few key points as follows:

Also, it is possible to combine this section with the introduction of tour optimisers (subsection optimisation in the tour)

huizezhang-sherry commented 3 years ago
  1. Gael also feels she doesn't know where to look in my graphs in the Visual Diagnostics section.

I explain these are sample plots I created and she needs to look at the Diagnosing an optimiser section to know what to look at in a specific projection problem. Gael thinks I need to make this clearer and I need to work on my figure captions to make them self-contained.

huizezhang-sherry commented 3 years ago
  1. Gael would like to see a discussion on the convergence of the optimisers

We could add this in the third and fourth dot point in the second issue

huizezhang-sherry commented 3 years ago
  1. Gael thinks the subtitle, tour looking at itself doesn't make much sense to her
huizezhang-sherry commented 3 years ago
  1. Gael suggests putting the algorithms in the appendix