Yesterday I was in Chicago for a full-day workshop on Bayesian statistics at Orbitz Worldwide. In the morning I presented “Learning to Love Bayesian Statistics”, an overview of Bayesian approaches and my attempt to debunk the myths. The attendees seemed interested, and asked great questions. Their photographer was kind enough to share some photos from the event; here’s a shot from the discussion that broke out after the talk:
Later I presented a customized version of my workshop, “Bayesian Statistics Made Simple.” Again, it was a great group of people with lots of excellent and challenging questions. As always, it’s interesting for me to hear about the problems people are working on and to apply data science tools to new challenges.
I was only in Chicago for 36 hours, but I got to walk around quite a bit, and enjoyed the city (and the perfect weather!).
This week I am visiting the University College at Twente, which is in the process of creating a new program in Technology and Liberal Arts & Sciences (ATLAS). When I arrived yesterday, they were just putting the new sign on the building:
Today I am teaching two workshops, one based on the Joy and Beauty of Computing curriculum, and one on Python.
Will Millennials Ever Get Married? Survival Analysis and Marriage Data
Recent studies report that an increasing share of Americans have never married, which suggests that current young adults might marry at lower rates than previous generations. Using data from a national survey, we find that successive generations are getting married later, but our predictions suggest that the fraction of people who eventually marry will not change substantially. Our analysis uses Pandas for data extraction and cleaning, bootstrap methods for working with stratified surveys, lifelines for survival analysis, and time series analysis with statsmodels. All code and data for this study is in a public repository.
Basic Sound Processing in Python
Digital signal processing (DSP) has applications in all areas of engineering and science, but DSP methods are not widely known. Python provides an opportunity to make DSP more accessible. In this talk, I present an introduction to DSP focused on sound-processing applications. I present tool for working with digital signals using NumPy, SciPy and IPython. Examples include spectral analysis of music, spectrograms, noise, filtering, and system characterization. This material is based on Think DSP, a work-in-progress book available at think-dsp.com.
It looks like there will be a lot of other great talks. I am looking forward to the conference in July!
One of the joys of working on free books is the chance to collaborate with people all over the world. The Python version of How to Think Like a Computer Scientist was translated by Jeff Elkner in 2000 (plus or minus), but I never met him in person until PyCon 2013!
This year, I finally met Charles “Dr. Chuck” Severance, who wrote Python for Informatics, which started as a modified version of How to Think, but has transmogrified into a substantially different new book, which Chuck uses in his Coursera online classes.
For the last day of Signals and Systems we had a design session where students and some visiting faculty made concept maps of the topics we covered this semester, then we redesigned the class, generating a lot of ideas for things we might do differently (better?) next time.
We heard a lot of thoughtful suggestions, so thank you to everyone who attended! If you are curious about the class, here’s the web page.
Comments Off on Last day of Signals and Systems 2015
Here’s the description for my Computational Statistics I:
Statistical inference is a fundamental tool in science and engineering, but it is often poorly understood. This tutorial uses computational methods, including Monte Carlo simulation and resampling, to explore estimation, hypothesis testing and statistical modeling. Attendees will develop understanding of statistical concepts and learn to use real data to answer relevant questions.
And Chris’s Computational Statistics II:
This intermediate-level tutorial will provide students with hands-on experience applying practical statistical modeling methods on real data. Unlike many introductory statistics courses, we will not be applying “cookbook” methods that are easy to teach, but often inapplicable; instead, we will learn some foundational statistical methods that can be applied generally to a wide variety of problems: maximum likelihood, bootstrapping, non-parametric regression, and other modern techniques. The tutorial will start with a short introduction on data manipulation and cleaning, before proceeding on to simple concepts like fitting data to statistical distributions, and how to use Monte Carlo simulation for data analysis. Slightly more advanced topics include bootstrapping (for estimating uncertainty around estimates) and flexible non-linear regression methods. By using and modifying hand-coded implementations of these techniques, students will gain an understanding of how each method works. Students will come away with knowledge of how to deal with very practical statistical problems, such as how to deal with missing data, how to check a statistical model for appropriateness, and how to properly express the uncertainty in the quantities estimated by statistical methods.
Participants can sign up for either tutorial, or both. Good seats still available!
Comments Off on Statistics workshops at SciPy 2015