Artificial intelligence, machine learning and convex optimisation

"Cutting plane" methods converge on the optimal values of a mathematical function by repeatedly cutting out regions of a much larger set of possibilities (gold sphere). Credit: Jose-Luis Olivares/MIT

Optimisation is everywhere

Source: Phys.org 23 October 2015

Optimisation problems are everywhere in engineering: Balancing design trade-offs is an optimisation problem, as are scheduling and logistical planning. The theory—and sometimes the implementation—of control systems relies heavily on optimisation, and so does machine learning, which has been the basis of most recent advances in artificial intelligence.

Last October, at the IEEE Symposium on Foundations of Computer Science, a trio of present and past MIT graduate students won a best-student-paper award for a new "cutting-plane" algorithm, a general-purpose algorithm for solving optimisation problems. The algorithm improves on the running time of its most efficient predecessor, and the researchers offer some reason to think that they may have reached the theoretical limit.

But they also present a new method for applying their general algorithm to specific problems, which yields huge efficiency gains—several orders of magnitude.

"What we are trying to do is revive people's interest in the general problem the algorithm solves," says Yin-Tat Lee, an MIT graduate student in mathematics and one of the paper's co-authors. "Previously, people needed to devise different algorithms for each problem, and then they needed to optimize them for a long time. Now we are saying, if for many problems, you have one algorithm, then, in practice, we can try to optimize over one algorithm instead of many algorithms, and we may have a better chance to get faster algorithms for many problems."

Lee is joined on the paper by Aaron Sidford, who was an MIT graduate student in electrical engineering and computer science when the work was done but is now at Microsoft Research New England, and by Sam Wong, who earned bachelor's and master's degrees in math and electrical engineering and computer science at MIT before moving to the University of California at Berkeley for his PhD.

Verdigris takes $9M to power its AI energy consumption analytics

Source: TechCrunch.com 24 March 2016

Pulling energy consumption data on large facilities and driving smarter and more responsive building operation — with the prize being major cost and resource savings.

b2b IoT startup Verdigris is playing in just such a space. The company took in a $6 million Series A round in December, which it’s just announcing now — and which includes, on top of that, a $3 million convertible seed, bringing its total raised to date to $9 million.

“The energy data is packaged securitized and sent in real-time to an enterprise grade server in the cloud powering our algorithms,” explains co-founder Mark Chung. “Features of the data are clustered through unsupervised Machine Learning into sets that reflect individual devices and device states.

“Once the clusters are learned, their representation is recorded in a library of devices. Convex optimisation techniques can then map the device representations to the incoming stream of data. If we don’t know what kind of equipment it is or it hasn’t been seen before, the AI will prompt the user or our professional services team for labelling information.”