Course talk:CPSC522/Recurrent Neural Networks

From UBC Wiki

Contents

Thread titleRepliesLast modified
Critique123:35, 8 February 2018
Critique 2123:34, 8 February 2018
Critique 3005:28, 8 February 2018

Not sure if we are supposed to mark already, either way here are my current comments:

Comments[wikitext]

I do not have to add much to this page. I am not an expert in this topic but the entry overall really well done. Here some minor nitpicks:

  • The solutions section and its discussion is really brief and might benefit from some elaboration. E.g., talk about the precise advantages and drawbacks of each aspect a little more. What do I gain or lose if I chose any of these optimisations? Or add some more references to the sections.
  • Figure 4 is a bit small and hard to read, I would enlarge it a bit more.
  • "Builds on" and "Related Pages": Is that all RNNs are related to, isn't there more to add? Don't RNNs also draw from Markov operations and various ML aspects? Or did you keep it short because this is all covered by the neural net page?
  • I think showing sample code of RNNs in Python or any other language could be great on this page. It may encourage people to go try and play with the code.
  • ReLU is not properly explained it seems, is it assumed to be background knowledge?
Marking Scheme[wikitext]

I a scale of 1 to 5, where 1 means "strongly disagree" and 5 means "strongly agree" please rate and comment on the following:

   The topic is relevant for the course. 5
   The writing is clear and the English is good. 5 
   The page is written at an appropriate level for CPSC 522 students (where the students have diverse backgrounds). 5
   The formalism (definitions, mathematics) was well chosen to make the page easier to understand. 5
   The abstract is a concise and clear summary. 5 (The abstract is quite short, but covers pretty much all the content in the page.)
   There were appropriate (original) examples that helped make the topic clear. 5 
   There was appropriate use of (pseudo-) code. 4 (I think the page might benefit from some more python code. )
   It had a good coverage of representations, semantics, inference and learning (as appropriate for the topic). 5
   It is correct. 5 (Did not see any glaring errors.)
   It was neither too short nor too long for the topic. 5 
   It was an appropriate unit for a page (it shouldn't be split into different topics or merged with another page). 5
   It links to appropriate other pages in the wiki. 5
   The references and links to external pages are well chosen. 5
   I would recommend this page to someone who wanted to find out about the topic. 4
   This page should be highlighted as an exemplary page for others to emulate. 5

If I was grading it out of 20, I would give it: 19

FabianNikolausTrutzRuffyVarga (talk)00:58, 6 February 2018

Thanks will incorporate suggested changes :)

KevinDsouza (talk)23:35, 8 February 2018
 

Critique 2

Comments I liked reading this informative article. I also loved the bibliography of this article and found it very relevant. IMHO, some questions that could be answered by the article to make it more insightful are as follows: Why can the functions in Recurrent networks be only a sigmoid or a tanh? How does RNN accommodate probability? BPTT also has the tendency of getting stuck in local minima. Why? Are there are any solutions to the problems with BPTT?


Nitpick:


Grading Scheme

   * The topic is relevant for the course: 5
   * The writing is clear and the English is good: 5
   * The page is written at an appropriate level for CPSC 522 students (where the students have diverse backgrounds): 5
   * The formalism (definitions, mathematics) was well chosen to make the page easier to understand: 5
   * The abstract is a concise and clear summary: 5
   * There were appropriate (original) examples that helped make the topic clear: 5
   * There was appropriate use of (pseudo-) code: 2
   * It had a good coverage of representations, semantics, inference and learning (as appropriate for the topic): 4
   * It is correct: 5
   * It was neither too short nor too long for the topic: 5
   * It was an appropriate unit for a page (it shouldn't be split into different topics or merged with another page): 5
   * It links to appropriate other pages in the wiki: 5
   * The references and links to external pages are well chosen: 5
   * I would recommend this page to someone who wanted to find out about the topic: 4
   * This page should be highlighted as an exemplary page for others to emulate: 4

If I was grading it out of 20, I would give it: 19

SurbhiAmeyaPalande (talk)08:13, 7 February 2018

Thank you for the suggestions :)

KevinDsouza (talk)23:34, 8 February 2018
 

Critique 3

Comments[wikitext]

I was very pleased with this page. I think it is well written and does a good job of exploring the topic. Well done!

Schema[wikitext]

The topic is relevant for the course. 5 The writing is clear and the English is good. 3 The page is written at an appropriate level for CPSC 522 students (where the students have diverse backgrounds). 5 The formalism (definitions, mathematics) was well chosen to make the page easier to understand. 4 The abstract is a concise and clear summary. 5 There were appropriate (original) examples that helped make the topic clear. 5 There was appropriate use of (pseudo-) code. 4 It had a good coverage of representations, semantics, inference and learning (as appropriate for the topic). 5 It is correct. 5 It was neither too short nor too long for the topic. 5 It was an appropriate unit for a page (it shouldn't be split into different topics or merged with another page). 5 It links to appropriate other pages in the wiki. 5 The references and links to external pages are well chosen. 5 I would recommend this page to someone who wanted to find out about the topic. 4 This page should be highlighted as an exemplary page for others to emulate. 5

If I was grading it out of 20, I would give it: 20

JocelynMinns (talk)05:27, 8 February 2018