Dimension reduction is the process of embedding high-dimensional data into a lower dimensional space to facilitate its analysis. In the Euclidean setting, one fundamental technique for dimension reduction is to apply a random linear map to the data. The question is how large the embedding dimension must be to ensure …
For stochastic control problems with uncountable state and action spaces, the computation of optimal policies is known to be prohibitively hard. In this talk, we will present conditions under which finite models obtained through quantization of the state and action sets can be used to construct approximately optimal policies. Under …
Environmental studies frequently involve understanding complex between interactive components which may be challenging to detect, quantify or analyze using traditional methods. Furthermore, the interactions between the components themselves might be challenging to model or empirically estimate. For example, a fundamental bottleneck to fingerprinting petroleum sampled from an oil-rich region lies …
We look at models of competition and risk within the context of power system markets. We demonstrate when social optima exist, what properties on risk measures and contracts are needed, and how to solve these problems in large scale practical settings. We look at specific structures that give rise to …
Speaker: Daniel Pimentel-Alarcon Title: Learning subspaces by pieces Abstract: We love subspaces. We observe a phenomenon and try to find a line that explains it. We get our hands on some data, and try to find a subspace that fits it. But in many relevant applications, data is missing, and …
Whither speech recognition? is the title of a well-known paper written by J. R. Pierce in 1969 which unfortunately stopped Bell Labs from continuing speech recognition research for several years at the beginning of the 1970s. Even though 45 years have passed since then and we have actually observed various …