We focus on a stochastic learning model where the learner observes a finite set of training examples and the output of the learning process is a data-dependent distribution over a space of hypotheses.

We focus on a stochastic learning model where the learner observes a finite set of training examples and the output of the learning process is a data-dependent distribution over a space of hypotheses. The learned data-dependent distribution is then used to make randomized predictions, and the high-level theme addressed here is guaranteeing the quality of predictions on examples that were not seen during training, i.e. generalization. In this setting the unknown quantity of interest is the expected risk of the data-dependent randomized predictor, for which upper bounds can be derived via a P AC-Bayes analysis, leading to P AC-Bayes bounds. Specifically, we present a basic P AC-Bayes inequality for stochastic kernels, from which one may derive extensions of various known P AC-Bayes bounds as well as novel bounds. We clarify the role of the requirements of fixed'data-free' priors, bounded losses, and i.i.d.

We'll expand the coverage of literature on P AC-Bayes bounds for non-i.i.d.

We introduce a novel topology, called Kernel Mean Embedding Topology, for stochastic kernels, in a weak and strong form. This topology, defined on the spaces of Bochner integrable functions from a signal space to a space of probability measures endowed with a Hilbert space structure, allows for a versatile formulation. This construction allows one to obtain both a strong and weak formulation. (i) For its weak formulation, we highlight the utility on relaxed policy spaces, and investigate connections with the Young narrow topology and Borkar (or $w^*$)-topology, and establish equivalence properties. We report that, while both the $w^*$-topology and kernel mean embedding topology are relatively compact, they are not closed. Conversely, while the Young narrow topology is closed, it lacks relative compactness. (ii) We show that the strong form provides an appropriate formulation for placing topologies on spaces of models characterized by stochastic kernels with explicit robustness and learning theoretic implications on optimal stochastic control under discounted or average cost criteria. (iii) We show that this topology possesses several properties making it ideal to study optimality, approximations, robustness and continuity properties. In particular, the kernel mean embedding topology has a Hilbert space structure, which is particularly useful for approximating stochastic kernels through simulation data.

We focus on a stochastic learning model where the learner observes a finite set of training examples and the output of the learning process is a data-dependent distribution over a space of hypotheses. The learned data-dependent distribution is then used to make randomized predictions, and the high-level theme addressed here is guaranteeing the quality of predictions on examples that were not seen during training, i.e. generalization. In this setting the unknown quantity of interest is the expected risk of the data-dependent randomized predictor, for which upper bounds can be derived via a PAC-Bayes analysis, leading to PAC-Bayes bounds. Specifically, we present a basic PAC-Bayes inequality for stochastic kernels, from which one may derive extensions of various known PAC-Bayes bounds as well as novel bounds. We clarify the role of the requirements of fixed 'data-free' priors, bounded losses, and i.i.d. data. We highlight that those requirements were used to upper-bound an exponential moment term, while the basic PAC-Bayes theorem remains valid without those restrictions. We present three bounds that illustrate the use of data-dependent priors, including one for the unbounded square loss.

Stochastic kernel based dimensionality reduction methods have become popular in the last decade. The central component of these methods is a symmetric kernel that quantifies the vicinity of pairs of data points and a kernel-induced Markov chain. Typically, the Markov chain is fully specified by the kernel through row normalization. However, it may be desirable to impose user-specified stationary-state and dynamical constraints on the Markov chain. Notably, no systematic framework exists to prescribe user-defined constraints on Markov chains. Here, we use a path entropy maximization based approach to derive Markov chains on data using a kernel and additional user-defined constraints. We illustrate the usefulness of the path entropy normalization procedure with multiple real and artificial data sets. All scripts are available at: https://github.com/dixitpd/maxcaldiffmap