These are my assignments for Introduction to Data Science - Bill Howe -- Spring 2013 (and again in Summery 2014), offered on Coursera.

Assignment 1: Parsing Twitter feeds. Uses python.

Assignment 2: Relational algebra and SQL. Uses SQLite3.

Assignment 3: Map/Reduce examples. Uses python.

Assignment 4: Optional Large-scale processing in The Cloud. Uses Pig/Map Reduce. You can run problems 1, 2a/b, and 3a/b on a local machine. p4 needs to be run on a cluster. The problem took approximately 20 minutes to run on an 18 XLarge2 Elastic MapReduce cluster, or US$12. Given that AWS bills by the hour, I would consider using fewer instances or the XLarge tier to optimize expenses.

Assignment 5: Machine learning with R using a real data set. This was a cool assignment because it had a degree of real-life (e.g., bad data)

Assignment 6: Optional visualization exercise using Tableau. For this, my dashboard was simply a graphic depicting the frequency, by state, of precautionary landings following an animal strike on a departure phase (run-up or climb). New York, Ohio, Illinois and Pennsylvania had the highes occurrence of events during climbing, Texas during run-up.

kaggle: Kaggle exercise. As this was listed after an optional "real-world" exercise (for which the choices were limited), I assumed it was optional. The Titanic example was too easy, so I opted for the Scikit-learn competorial.