It’s an elegant idea I haven’t run into before: gather data on site preferences by selecting what version to present on the epsilon-greedy solution to the multi-armed bandit problem and just letting it run. You’re looking at a setting where effectiveness can be easily measured, such as by clickthrough, but the contrast is with A/B testing where the effect of a single change is being measured for a time and then a switch is being made, if desirable. Comments suggest tweaks/details like ensuring that a single visitor sees a consistent view of the site, at least for small windows of … Continue reading Machine Learning in Usability Testing
Draw a Stickman is just delightful. Go to that page, select Episode 1, and I defy you not to say “Oh cool!” within 20 seconds. It is entirely charming. What else could you want on a Monday afternoon :)
Things my RSS feed wants me to do this week: Stay productive after work on my side projects because if work and your homelife are all you’re doing, you’re a bum. Go to NYC for Manhattanhenge or make plans to go back for it in July. Buy Travel Blog the Board Game – even though from the description it is unclear where the “Blog” part of the game comes in. Consider if I am suprised that those over 55 pick more secure passwords than those under 25.
I’ve been learning some Octave recently and have refreshing my Python on my summer to-do list for a course I’ll be teaching in the fall, plus I’ve been running into a ton of articles about R (particularly for data visualization) that are making me think I ought to give it a look as well. So this comparison of the three from Slashdot was a nice overview from one person’s experience of which tool to turn to when: R, Octave, and Python: Which Suits Your Analysis Needs?. The comments (as always) offer some interesting input as well, including suggestions for other … Continue reading Choosing an Analysis Tool