The Ellsberg paradox is a puzzle concerning human rationality. It postulates a scenario involving an urn with 90 balls inside - 30 red ones, while the remaining 60 can be yellow or black. A player gets to draw a ball from the urn, and receives a payoff dependent on the color …

A book on Procedural Content Generation in Games. Very nice.

The Myth of Schemaless.

A classic from the MIT AI lab - Hakmem.

Less research is needed.

Cache coherency primer, about cpu caching.

Favor Disjuction over Validation when handling errors in Scala.

A Lispers first impression of Julia. I'm not a …

Suppose you have a method of making decisions that doesn't suck - i.e., one that is objective, consistent, and never leaves you guessing about what to do. The mathematical field of topology provides a wonderful theorem - if you have a good decision process, it is equivalent to managing by metric.

Mapping sets in Scala using Coyoneda

Why Dependently Typed Programming Will (One Day) Rock Your World. Related thoughts on type errors in mathematics education.

Constructions vs Specifications.

How to use Treelog, and the underlying github page.

Frequentist inference only seems easy.

A stochastic diffusion process for the Dirichlet distribution. Some …

Strength is a category theoretical property which is known to be boring - every Monad in Hask (and similarly Scal) satisfies it. The question arises - what about it's dual, Costrength? What would a Costrong Comonad look like? Is it interesting? It turns out the answer is no - every Comonad in Hask/Scal is Costrong. In this post I'll provide a brief tutorial on Comonads and show why every Comonad is also Costrong. You do NOT need to know much category theory to follow this post, but some familiarity with the everyday programming of Monads will be helpful.

I was recently writing some Scala code, and I wanted to return an array of generic type:

class Foo[+T] {
  def someArray(size: Int): Array[T] = {
    val result = new Array[T](size)
    ...
    result
  }
}

This did not make Scala happy:

[info] Compiling 1 Scala source to /home/stucchio/src/breeze/math …

I'm currently dating two lovely women, Svetlana and Elise. Unfortunately continuing to date both of them is unsustainable so I must choose one.

In order to make such a choice, I wish to construct a ranking function - a function which takes as input the characteristics of a woman and returns …

Tyler Cowen explains how to eat well. Read this one.

Great blog post showing the difference between confidence and credible intervals.

The Safyness of Static Typing.

Get the Science out of your Yogurt

Gelman, Hill and Yajima on Bayesian multiple comparisons via multilevel models.

Once in a while Michael O …

Earlier this year I published a blog post about a Baysian decision rule for choosing between two variations, each with a potentially different conversion rate. Later Evan Miller wrote a blog post computing an exact formula for evaluating it, rather than approximating the integral numerically. In this post I'm going …

Earlier this year I published a blog post about a Baysian decision rule (now dead link, but VWO now uses this for A/B testing and tech docs are here) for choosing between two variations, each with a potentially different conversion rate. The basic idea of the decision rule is …