Further features of the hat matrix; y=Hy, X=HX, e=(I-H)y [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Properties of (I-H); symmetric, idempotent, trace = n-p [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
This video shows that the Binomial belongs to the Exponential Family. [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
In this video Bayes' theorem is applied to the Binomial density, choosing an appropriate conjugate prior. [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Brief slide explaining how to estimate phi in a GLM [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
The Normal density written in (Canonical) form of the Exponential Family format suitable for GLMs [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Brief outline of the canonical form of the exponential family, illustrated with the Poisson distribution. [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
This briefly and roughly introduces the standard uniform distribution [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Here we set out the simplest statistical model we could perhaps suggest for the regression situation, and see what beta_2 is in terms of moments [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Maximum likelihood for the Normal distribution [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
This attempts to sketch out regression as a problem for vec(y), vec(x1), vec(x2) etc. [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
A brief introduction to the hat matrix, and some ideas around leverage [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Assumes we've done this in a previous course (linear algebra/calculus or even statistics) [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
We briefly outline the kind of work we intend to develop by examining a simple regression model fitted to the Zagat data on New York Restaurants. [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
A description of the Moment Generating Function for the Poisson distribution, and one example of its use. [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Brief explanation of the Poisson probability mass function (no explanation of where the formula comes from, but there is a check that it is a valid probability function and we find E[X]) [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Description of the Binomial(n,p) distribution, find the first moment, state Var(X) [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Some properties and derivation of the Geometric distribution, including check on validity and finding E[X] [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
A brief description of how we obtain the Binomial m.l.e. [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]
Finds the maximum likelihood estimator for the exponential distribution [Released under Creative Commons Attribution-NonCommercial-ShareAlike 2.0 UK: England & Wales licence.]