Applications of approximate inference techniques for optimal design in self-assembly and automated programming

March 22, 2006
Vladimir Jojic | University of Toronto

In this talk I will describe two design problems in areas of chemistry and computer science which yield themselves to machine learning techniques.

The area of supra-molecular chemistry deals with mechanisms of noncovalent assembly of particles. The interest in understanding the underlying mechanisms is two-fold: it provides insight into protein complex formation and paves the way for application of these mechanisms in nanotechnology. The spontaneous assembly of particles is referred to as self-assembly and the ability to design such processes holds great promise for nanotechnology. The problem of design of self-assembly processes turns out to be a task surprisingly familiar to the machine learning community: that of probability maximization. The standard Boltzmann machine learning rule can be applied to this task, and I will demonstrate a way to speed up the evaluation of the derivatives via importance sampling. In addition, I will demonstrate several methods for evaluating the probability of a shape under a self-assembly process.

The second part of my talk will be devoted to another problem which at first blush does not admit a probabilistic interpretation. This is the problem of inferring programs given input/output pairs. I will introduce a probabilistic representation of code. This representation gives rise to a distribution of state sequences and allows approximate inference in form of loopy belief propagation. I will illustrate the performance of this method on tasks of discovering programs for polynomial computation and list reversal given only examples of the input/output pairs.

Speaker Details

I am a PhD candidate in Computer Science at University of Toronto. Prior to graduate school, I worked at Microsoft as an SDE in ASP.NET group and at the Santa Fe Institute. Since 2004 I have been a Microsoft Fellow. I enjoy devising approximate inference techniques for difficult problems such as the two above. I spend most of my efforts in computational biology and occasionally cross over into computational chemistry in hopes that molecular dynamics can yield efficient answers to biological questions.