Steven C. Howell, PhD
Data Scientist at Booz Allen Hamilton
howell_steven@bah.com
@StevenCHowell
Lexington Park, MD
Say you have a bag of 99 fair coins and 1 unfair coin that has heads on both sides. You draw a coin from the bag and flip it 10 times, each time showing heads.
What is the probability you drew the unfair coin?
This talk will illustrate how to simulate what seems like a complex statistics problem using your Python hacking skills.
Inspired by Jake Vanderplas' PyCon 2016 talk "Statistics for Hacker".
Statistics can often be unintuitive, and even frustrating to wade through boring equations. Can I use my ability to program to hack my way to a solution?
Computers are great at simulating complex situations. If we can define a system, the computer can quickly iterate through thousands or millions of possibilities to get to converge on an answer. In this talk, we will see how to to use hacking skills to simulate real systems. Additionally, we will be able to watch a live-streamed plot of the results as the computer converges on a solution.
Outline 10 min: Introduce the problem, Bayes' Theorem, the calculate the analytic solution. 15 min: Demonstrating how we can simulate the problem in Python. 5 min: Demonstrating a live-streamed plot analyzing the results.