|
|
|
|
|
|
by tomnipotent
4035 days ago
|
|
|
It's hard to take your reply seriously as it comes across like a pissing contest. Post is from 2005, and the author is Andrew W. Moore - Dean of CompSci at Carnegie Mellon and previously a VP Engineering at Google. But by all means, continue pissing while the rest of us appreciate the free content made available to us. |
|
|
But I'm judging Moore's materials based on the materials, not his employment history.
I have nothing against Moore; it's not about Moore or me. Instead, it's about what Moore wrote.
Google, CMU CS aside, sorry to tell you, or maybe it's good news, think of the good news, instead of that A. Moore material, there is much, much higher quality material going way back, e.g., already by, say, 1970. There's G. Dantzig, R. Gomory, R. Bellman, G. Nemhauser, Ford and Fulkerson, P. Wolfe (e.g., Wolfe dual in quadratic programming), R. Bixby, H. Kuhn, A. Tucker (prof of the prof that was the Chair of my Graduate Board orals), D. Bertsekas, J. von Neumann, J. Nash, R. Rockafellar, W. Cunningham, and many, many more. None of these people is in computer science.
E.g., there are stacks of books on multivariate statistics with linear discriminate analysis; there's log-linear for categorical data analysis; there's controlled Markov processes and continuous time stochastic optimal control, with, say, measurable selection, scenario aggregation. etc.; there's lots of material on resampling, the bootstrap (I have published a paper in essentially that topic); there's sufficient statistics from the Radon-Nikodym theorem; and much more.
Okay, just in regression and multi-variate statistics, just from my bookshelf, there's:
William W. Cooley and Paul R. Lohnes, 'Multivariate Data Analysis', John Wiley and Sons, New York, 1971.
Maurice M. Tatsuoka, 'Multivariate Analysis: Techniques for Educational and Psychological Research', John Wiley and Sons, 1971.
C. Radhakrishna Rao, 'Linear Statistical Inference and Its Applications: Second Edition', ISBN 0-471-70823-2, John Wiley and Sons, New York, 1967.
N. R. Draper and H. Smith, 'Applied Regression Analysis', John Wiley and Sons, New York, 1968.
Leo Breiman, Jerome H. Friedman, Richard A. Olshen, Charles J. Stone, 'Classification and Regression Trees', ISBN 0-534-98054-6, Wadsworth & Brooks/Cole, Pacific Grove, California, 1984.
And their mathematical notation is quite precise.
In simple terms, what Moore is doing in those notes is mostly, not all, some very fine, old wine, corrupted, and in new bottles with new labels. E.g., the 22 pages on game theory without mentioning just the simple linear programming solution is gentleman D- work.
Readers would be seriously mislead and ill-served not to hear that the Moore material is inferior, really, not good. We're talking grade C, gentleman B-.
Think of the good news: There's much, much better material long on the shelves of the research libraries although rarely as computer science.
That's just the way it is. People here on HN should be aware of that situation.