Course talk:CPSC522/Transfer Learning with Markov Logic

Transfer Learning with Markov Logic

Comments

This is a really good page, I do not have any major complaints. The structure is very good in terms of framing and juxtaposing the papers. Everything is clear and understandable. I am missing a bit of background on transfer learning, the Wikipedia reference seems a bit weak. But that is okay since it is considered background. How does transfer learning related to deep reinforcement learning that is model free? E.g. AlphaGo Zero, which can also play other games. Does transfer learning skip the training stage entirely? How about adding a concrete example for the DTM and TransferNew, so that one can see the concrete steps of the algorithm in more detail?

Minor things: Kok and Domingos -> should be cited "Put links and content here to be added. This does not need to be organized, and will not be graded as part of the page. If you find something that might be useful for a page, feel free to put it here." -> this can be cut?

Marking Scheme

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 
   There were appropriate (original) examples that helped make the topic clear. 3
   There was appropriate use of (pseudo-) code. 5
   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. 5
   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

Thank you for the feedback. I have made the link to the Transfer Learning paper more explicit in the Background if the reader wants more information. Kok and Domingos has already been cited. I've also removed the default text in "To Add". It is difficult to give concrete steps of both algorithms as they are not provided in either paper, but I've added a few examples of the theory and intuition behind them.

This is a really well done page, properly organized and straight to the point on an interesting topic. On the second paper I think a brief example of two target domains with shared second-order structure would be helpful for understanding.

On 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 it's complex but the background on MLNs is also on the wiki
• The formalism (definitions, mathematics) was well chosen to make the page easier to understand. 4 some of the formalisms could use additional explanation, e.g. on homophily
• The abstract is a concise and clear summary. 4 maybe a little too concise!
• There were appropriate (original) examples that helped make the topic clear. 3 I think there are spots examples would be useful here
• There was appropriate use of (pseudo-) code. NA
• 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 MLNs already separate
• 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. 5
• 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

Thank you for the critique. I've added a little more to the abstract, included a few examples, and added a link to the homophily wikipedia article. There's also an example on homophily with example predicates.

• 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. 5
• 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.
• 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. 5
• 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

I'm very pleased with this page. Well written and outlines the papers efficiently. I have nothing to add.

Thank you for the critique.