---

Probability and Statistics for Computer Science
STAT 391 Spring Quarter 2021

---

Home Course Description Syllabus Books Class mailing list Anonymous feedback Assignments Handouts Resources UW Statistics UW ACMS Program UW CSE

Resources

Probability and statistics resources The Books page The Statistics Tutoring Center The excellent course notes for CSE 312 (pre-2020 web pages may have more openly available notes) Resources for the Boostrap, Confidence sets and Linear Regression All of statistics by Larry Wasserman Linear regression pp 245 The bootstrap pp 129 Confidence sets pp 116 (534) All of non-parametric statistics by Larry Wasserman (Bootstrap Ch 3, Cross-Validation 6.1 pp 125) Basic introduction to linear regression by David Lane Introduction to linear regression by Rencher and Schaalie (Linear Models in Statistics, Chapter 6, page 127/139). Another intro to linear regression, this time directly with multivariate data (The Elements of Statistical Learning, Chapter 2.3 pp30, and Chapter 3.2, pp 44-47) by Hastie, Tibshirani and Friedman Notes on logistic regression and naive Bayes by Tom Mitchell an illustration of linear regression from Hastie, Tibshirani and Friedman Boston Housing Data example Python resources (compilation and suggestions by Hoyt Koepke) A General Tutorial can be found at http://docs.python.org/2/tutorial/. API and library reference: http://docs.python.org/2/py-modindex.html. Numpy Tutorial: http://www.scipy.org/Tentative_NumPy_Tutorial. Numpy Reference: http://docs.scipy.org/doc/numpy/reference/. Scipy Reference: http://docs.scipy.org/doc/scipy/reference/. Matplotlib Tutorial: http://matplotlib.org/users/pyplot_tutorial.html. List of matplotlib examples: http://matplotlib.org/examples/pylab_examples/index.html. Matplotlib Gallery (browse examples by picture): http://matplotlib.org/gallery.html#pylab_examples Installation, etc: If you run Linux, the needed libraries are all in the standard repositories. At a minimum, you will need NumPy, SciPy, and Matplotlib. An easy way to get all the packages you want is is with Miniconda. The open source IDE Spyder provides a decent editor with an interactive command prompt designed for writing scientific code. Other possible options include Wingware and tons of others. A bunch of results are here: http://stackoverflow.com/questions/81584/what-ide-to-use-for-python. Feel free to post to the forum thoughts about editors. For other operating systems, I’d strongly recommend using a pre-packaged installer for Python with all the needed scientific libraries such as Pythonxy or the Enthought Python Distribution (you’ll need to get a free academic license). Matlab resources This is a Matlab tutorial script file which accompanies the notes on how to make plots with Matlab. Also there is more complete introduction to Matlab here. and complete help at http://www.mathworks.com Calculus tutorial Calculus Tutorial in Harvey Mudd College Mathematics Department Online tutorials from Hofstra University Calculus.org: The calculus page has a lot of useful links for calculus tutorial and help Pretty applications of probability and statistics in computer science and related fields: Error correcting probabilistic decoding the best error correcting codes in existence use probabilistic reasoning Retina registration: An example in computer vision Covid-19 predictions. How accurate were the models? This paper shows the models were not so accurate. But think also of the quality of data the data scientists had to work with. And another research paper with more detail on the models used. A forest of sensors: automatic camera surveillance A natural language question answering system; how do you like the way it works? and Robot localization with particle filters Other (more or less related) resources Netica a Bayes nets simulation system (free downloads of demo version) The vocal joystick project, Jeff Bilmes', UW EE Stochastic motion textures GRAIL (the Graphics and Image Laboratory), UW The www.Kernel-Machines.org resource page (last updated 2008 but still good) Your suggestion here