Digging into Data

• Williams, Chapter 1 - 2.5
• Librarians: Get a Job in Data Science, Make Lots of Money
• The Age of Big Data

• R for OS X
• R for Windows
• Weather data
• RStudio (optional)

• A probability primer
• (Optional, if you want more detail) Grinstead and Snell, Chapters 1-2, Chapter 4.1
• In class exercies

• Williams 3-3.5, 5
• Skim Williams 4 (Know what's there so you can use it as a reference later)
• (Optional) ggplot2 tutorial

• CaBi ridership data
• XML list of CaBi stations
• How to import XML into a spreadsheet
• Useful Google spreadsheet formulas
• A nice introduction to the R commands I'll use outside Rattle

• Google spreadsheet
• input csv (downloaded from Google spreadsheets)
• Python script
• Python output
• Faceted output from ggplot2 (other outputs)
• R Script
• R output csv

• Handout
• R Commands

• Williams 7
• Chapter 3 of Gelman and Hill (to be handed out in class). Focus on the concepts, not the R commands (unless you want to).
• (Optional, if you want more detail) Andrew Ng's regression notes
• (Option for review of normal distribution) The Normal Distribution

• HW1 Preview
• Linear Regression Handout
• Regularized Linear Regression Solution

• Williams 8, 15
• Naive Bayes Tutorial (Stop when you get to the hands on example)
• Tom Mitchell's Math-heavy introduction to NB and LR (optional)
• Ng and Jordan (only read this if you thought that the Mitchell reading wasn't theoretical enough)

• Classification Handout

• Williams 11, 14, 16

• Decision Tree / SVM Handout

If you feel you have a good handle on the content, you don't have to come to the review. Please do, however, at least look at the practice problems. If you can complete the problems without any difficulty, then feel free to skip the class (get started thinking about your project!).

The midterm will be in-class. There will be 20-50 multiple choice questions and 1-3 short answer questions. The questions will be similar to the in class exercises and review questions. There will be example questions provided before the midterm review.

You can bring a simple calculator (no device with an internet connection, even if it's disabled) and one A4 or US letter sheet of paper with notes (front and back). Your note sheet need not be hand-written, but you should prepare it yourself (it's a useful exercise).

• Williams 15
• Resnik and Lin's Evaluation in NLP
• Publication that resulted from students' work on feature engineering homework in this class.

• Part of Speech Dataset
• Evaluating features on Part of Speech

• Williams 9
• Probabilistic Topic Models
• Topics in Novels (Optional)
• LDA package for R (Optional)
• Code that does K-means clustering

• Closing the Loop: Fast, Interactive Semi-Supervised Annotation With Queries on Features and Instances

• ConvNet JS