Columbia

Analytics graduate course.  Diary of additional information and resources.  Bookmark this permanent site and similar to Canvas course system, and please check in every so often.  Thanks!

2017:

The Statistics Topics book is on sale the first week of the course (one dollar donation or free for Kindle and Prime users): https://www.amazon.com/Statistics-Topics-Salil-Mehta/dp/1499273533 and https://sites.google.com/site/statisticalideas/statistics-topics

Also our newest color interactive product coming late 2019. Enjoy the teaser trailer: here.

Please sign-up for the free web log for ideas on crafting probability and statistics research, and access to some free public data sets (must accept confirmation e-mail): http://feedburner.google.com/fb/a/mailverify?uri=StatisticalIdeas and https://twitter.com/salilstatistics

The dataset for your project should be at least 1,000 records to have some meat on the analysis tools, but perhaps sampled down if greater than 100,000.

You should be thinking about your data and type of analysis today and every day until you submit on Class 3 following the canvas and in-class instructions.  The typical analysis could be a predictive regression, clusters, or a probability analysis on a market basket. 

You might want to download into a special folder all of the datasets for the course so that you can readily access them, without internet access, at your leisure

Please see current timeline of deliverables (note that future deliverables in the 2nd half of the class will change slightly as the course progresses, depending on a variety of needs)

Notes from last year, September 12:

Keep in mind from our Target case study that the search for a potentially new and powerful customer base drove their analytics, not analytics for analytics sake.

Also keep in mind that only Target customers who they have information on can be scored, not say customers who browse Walmart.

Notes from last year, September 17:

Recall that for the time being class will begin at 9:15am given the current train schedule on Saturday mornings.  Make sure your Canvas works this week.  I am not allowed to accept and grade assignments delivered to me directly (outside of an auditable Canvas system).

Finally, there are some additional interesting articles and resources that you might enjoy.  I show them here (this list below will evolve over the course of this and future semester to focus on a smart variety of places to explore). 

Sample of data analysis (my research generally includes the underlying data, others do not):

• http://www.nytimes.com/interactive/2016/09/08/us/us-murder-rates.html

• http://www.nytimes.com/2014/04/06/business/the-oracle-of-omaha-lately-looking-a-bit-ordinary.html

• https://www.washingtonpost.com/news/wonk/wp/2016/08/30/the-sneaky-math-that-made-the-lottery-more-alluring-and-harder-to-win/

• http://www.marketwatch.com/story/investors-are-better-off-flipping-a-coin-than-following-wall-street-pros-data-shows-2016-01-20

• http://www.zerohedge.com/news/2016-09-10/deutsche-warns-10-decline-market-reaches-mania-level

• https://mishtalk.com/2016/04/14/silver-vs-mish/

• http://statisticalideas.blogspot.com/2015/05/a-face-in-numbers.html

• http://statisticalideas.blogspot.com/2015/07/volatility-in-motion.html

• https://www.whitehouse.gov/administration/eop/cea/economic-report-of-the-President/2016

• http://science.sciencemag.org/content/343/6176/1203 or http://gking.harvard.edu/files/gking/files/0314policyforumff.pdf

• http://www.nytimes.com/2012/02/19/magazine/shopping-habits.html

• http://www.nature.com/news/scientific-method-statistical-errors-1.14700

• http://www.nytimes.com/2016/07/13/business/tesla-autopilot-fatal-crash-investigation.html

• http://statisticalideas.blogspot.com/2016/08/indias-olympic-triumph.html

Sample of data and visual pools:

• http://nber.org/cycles/main.html

• https://www.treasury.gov/initiatives/financial-stability/reports/Pages/TARP-Tracker.aspx

• http://www.cancerresearchuk.org/health-professional/cancer-statistics/survival

• https://data.nasdaq.com/

• https://www.theguardian.com/us-news/series/counted-us-police-killings

• http://www.bls.gov/data/

• http://apps.who.int/gho/data/node.home

• https://www.ncdc.noaa.gov/data-access/model-data/model-datasets

• http://www.bloomberg.com/graphics/infographics/

• https://www.fbi.gov/stats-services/publications

• http://www.fda.gov/iceci/enforcementactions/warningletters/2013/ucm376296.htm

end of message 

NOTE, THE 2016 FILES BELOW DO NOT APPLY FOR 2017 COURSE