I continue to send gentle nudges about the usefulness of bandit algorithms, to quote “By using a bandit approach you can determine if say option A is currently better (w.r.t. some confidence bounds). After each visit you can update the bounds and after a sufficient number visits you have a winner. The main difference to more traditional AB testing is that the process is more adaptive and less time is wasted on exposing an inferior product to the user.” Imagine my delight when I saw this morning a free draft copy of an extensive book (PDF) on the subject, by Tor Lattimore and Csaba Szepesvári!
I also enjoyed two YouTube lectures over the last couple of weeks:
- If you’ve ever wondered what “Tensor” really means see Tamara Kolda’s excellent lecture.
- This London PyData talk by Vincent D. Warmerdam has some nice machine learning tricks and tips.
* * *
Graphical comparison between model predictions and real-life observations is a challenging task. The problem for continuous values is a relatively easy one — you create a scatter plot with the predicted values on the X-axis and observed values on the Y-axis (not the other way around! See this paper to understand why.
Until recently, I couldn’t find a useful graph that would compare predictions and observations for a binary label case. Last week, I stumbled on a 2011 paper by Greenhill et al titled “The Separation Plots: A New Visual Method for Evaluating the Fit of Binary Models.” In that paper, the authors propose a very dense and informative way to look at the results of a binary classification problem.
* * *
In a piece I recently highlighted on on our sister site Longreads, Polina Aronson and Judith Duportail examine the cultural implications of bots and artificial intelligence. People don’t just ask Siri to convert from metric to imperial measures — they sometimes share their deepest feelings. Aronson and Duportail offer a fascinating look at the cultural foundations behind the responses that artificial intelligence like Siri and Alisa offer to expressions of human feeling and emotion.
What reads have recently rocked your world in the field of Data Science? Leave us a comment!