Spring 2014: Meetings 4-6
Thu 29 May 2014 by Nate Arnett, Karl DeVries, Adrian BrasoveanuThe 6th and final meeting of the quarter will take place next Thur, June 5 (the last week of classes). Meetings 4-5 (May 22, and May 29) discussed Bayesian inference in more detail (e.g., how to do model comparison with Bayes factors), and introduced MCMC estimation and the way JAGS enables us to do MCMC based on little more than a specification of the probabilistic model we assume generated the data. We started by estimating posterior distributions for very simple models with both continuous (normally-distributed) responses and categorical (Bernoulli-distributed) responses.
The last meeting of the quarter will be dedicated to a discussion of Bayesian inference for more complex / realistic linear models.
We will probably continue with the 2 books mentioned in the previous post (Bayesian Cognitive Modeling: A Practical Course, Lee & Wagenmakers 2014, and Bayesian Rationality: The Probabilistic Approach to Human Reasoning, Oaksford & Chater 2007) over the summer.
Spring 2014: Meetings 1-3
The plan for the spring quarter and possibly early summer is to read and work through the following two books:
- Bayesian Cognitive Modeling: A Practical Course (Lee & Wagenmakers 2014): see this pdf for the first two parts of the book and also this website for additional materials (code, answers to …
Winter 2014: Meetings 5-7
The 5th meeting this quarter was dedicated to a discussion of Hale (2011), led by Karl. His handout is available here.
The 6th meeting was dedicated to a discussion of NL-Soar led by Nate and based primarily on Lewis’s ‘93 dissertation: An Architecturally-based Theory of Human Sentence Comprehension …
read moreWinter 2014: Meeting 4
The fourth meeting this quarter took place on Fri Feb. 7. There won’t be any meeting on Fri Feb. 14, our next meeting will be on Fri Feb. 21 (same time and place) and Karl will lead the discussion of Hale’s 2011 paper. Nate will then guide us …
read moreWinter 2014: Meeting 3
The third meeting this quarter took place on Fri Jan. 31. We discussed garden-path phenomena based on ch. 4 of Hale’s ms. and the general view of parsing (and other language-related cognitive processes, e.g., learning) as heuristic-based search through an appropriate state space. We discussed (greedy) best-first and …
read moreWinter 2014: Meeting 2
The second meeting this quarter took place on Fri Jan. 24. We looked at some NLTK parsing demos, namely recursive descent (top-down, depth-first), bottom-up and left-corner parsers.
Nate discussed left-corner parsing in detail and the empirical evidence that favors left-corner parsers as models of the human parser / processor over purely …
read moreWinter 2014: Meeting 1
The first LaLoCo meeting this quarter took place on Fri Jan. 17 and it was well attended — thank you everyone for coming and participating!
We discussed chapters 3 and 4 of John Hale’s textbook draft, in particular: regular and context free grammars, finite state and push down automata, and …
read moreWinter 2014: Plan
The first LaLoCo meeting this quarter will take place on Fri Jan. 17 — not this coming Fri, but the next one. Regular time: Fridays, 2:15-3:15 pm. Place: Meeting & Analysis room (Stevenson 217).
Karl, Nate and I will do a series of lab meetings about parsing this quarter. The …
read more