Archive for the ‘argonauts’ Category

Hypothesis testing

Friday, June 23rd, 2006

It’s been almost 2 months since the last meeting, but this term has been a complete mess. Also, forcing people into slavery to have them learn about Clinical Statistics is proving harder than I initially thought.

But at last we had a look at Hypothesis Testing. And it was a meeting that left me totally knackered, in part because we have changed the format and this one was a bit more like imparting a Seminar, rather than a typical each-man-for-himself.

Argonauts: Hypothesis testing


Introduction to Clinical Statistics

Sunday, April 30th, 2006

The Argonauts become the R-go-nuts this Trinity Term, and we are going to study Clinical Statistics and GNU R. So today we had an Introduction to Clinical Statistics.


The Expectation-Maximization algorithm (II)

Sunday, April 16th, 2006

We had a second go at the EM-algorithm, after the previous fiasco. Well, this meeting went very well!

The minutes of the Expectation-Maximization (EM) algorithm are available on the Argonauts website, as usual.


The Expectation-Maximization algorithm

Sunday, March 19th, 2006

Meeting of the Argonauts to learn about the Expectation-Maximization (EM) algorithm. Absolutely pants. Worst meeting ever. We didn’t get anywhere, we got lost into details so we will have to repeat it again, if possible with more people.


MRFs in a Bayesian framework

Saturday, March 11th, 2006

Meeting of the Argonauts. We saw how to use Markov Random Fiels to calculate the priors in a Bayesian framework. Um, we didn’t quite finish it, but all this is very interesting and suddenly many things that we have seen before, like Kalman Filters come together.


Maximum a posteriori

Sunday, February 26th, 2006

Argonauts meeting today from 10.30am to 1.15pm. We saw Bayes’ Theorem, and Maximum Likelihood (ML) and Maximum a Posteriori (MAP) estimates, a way to combine information from different sources.


Self-Organizing Maps

Saturday, February 18th, 2006

Argonauts meeting today from 11am to 1pm. We did Self-Organizing Maps (SOMs), a way to visualize high-dimensional data using a neural network devised by T. Kohonen in 1982.