• New Kaggle competition: How Much Did It Rain? II.

  • I've just discovered that FiveThirtyEight has a wonderful daily newsletter: Significant Digits.

  • If you use R, you might have had your not always pleasant encounters with the stringsAsFactors option in functions like read.csv. This article explains why this has been done that way, and why this is set to TRUE by default.

  • Bagging and boosting are two different concepts that sometimes get mixed. Here you have a wonderful presentation that will help you understand what they are, once and for all.

  • In science, data sharing is a crucial step. We've had a very good example of this this week. From the article: A medical journal criticised British drugmaker GlaxoSmithKline on Thursday for delaying access to key data from a trial of its antidepressant paroxetine (Seroxat, Paxil) that would have shown earlier that it is neither safe or effective in adolescents.