McNulty // w5d2

Winter 2015
02/10/2015

Planned schedule and activities

9:00 am: Dawn of a new day

9:15 am: Categorical Data Analysis, Bernoulli and Binomial Distributions, Likelihood

10:15 am: Work on McNulty

11:00 am: Speaker -- Lee Madoff from Ernst Young

12:00pm: Food

1:30pm: Naive Bayes Classification

2:10pm: Work on McNulty

5:00pm: End of a beautiful day

Lecture Notes

Naive Bayes Slides

w5d2_Digits_Naive_Bayes.ipynb (52.7 KB)

Reading

Simple Naive Bayes Example
Another clear Naive Bayes explanation with distribution math
Naive Bayes in Wikipedia
Maximum Likelihood
Why does naive Bayes work so well even though the assumption is very bad?
Paper on spam filtering and types of naive Bayes (Gaussian, Multinomial, etc)
More technical paper with math on naive Bayes
Bayesian Inference (Think Bayes is also very good at explaining this)