Inverse reinforcement learning (IRL) aims to recover the reward function underlying a Markov Decision Process from behaviors of experts in support of decision-making. Most recent work on IRL assumes the same level of trustworthiness of all expert behaviors, and frames IRL as a process of seeking reward function that makes those behaviors appear (near)-optimal. However, it is common in reality that noisy expert behaviors disobeying the optimal policy exist, which may degrade the IRL performance significantly. To address this issue, in this paper, we develop a robust IRL framework that can accurately estimate the reward function in the presence of behavior noise. In particular, we focus on a special type of behavior noise referred to as sparse noise due to its wide popularity in real-world behavior data. To model such noise, we introduce a novel latent variable characterizing the reliability of each expert action and use Laplace distribution as its prior. We then devise an EM algorithm with a novel variational inference procedure in the E-step, which can automatically identify and remove behavior noise in reward learning. Experiments on both synthetic data and real vehicle routing data with noticeable behavior noise show significant improvement of our method over previous approaches in learning accuracy, and also show its power in de-noising behavior data.

Infinite SVM (iSVM) is a Dirichlet process (DP) mixture of large-margin classifiers. Though flexible in learning nonlinear classifiers and discovering latent clustering structures, iSVM has a difficult inference task and existing methods could hinder its applicability to large-scale problems. This paper presents a small-variance asymptotic analysis to derive a simple and efficient algorithm, which monotonically optimizes a max-margin DP-means (M2DPM) problem, an extension of DP-means for both predictive learning and descriptive clustering. Our analysis is built on Gibbs infinite SVMs, an alternative DP mixture of large-margin machines, which admits a partially collapsed Gibbs sampler without truncation by exploring data augmentation techniques. Experimental results show that M2DPM runs much faster than similar algorithms without sacrificing prediction accuracies.

In this paper, we present a hybrid grammar formalism designed to learn structured models of natural iconic gesture performances that allow for compressed representation and robust recognition. We analyze a dataset of iconic gestures and show how the proposed Feature-based Stochastic Context-Free Grammar (FSCFG) can generalize over both structural and feature-based variations among different gesture performances.

Policy gradient reinforcement learning (PGRL) has been receiving substantial attention as a mean for seeking stochastic policies that maximize cumulative reward. However, the learning speed of PGRL is known to decrease substantially when PGRL explores the policies that give the Markov chains having long mixing time. We study a new approach of regularizing how the PGRL explores the policies by the use of the hitting time of the Markov chains. The hitting time gives an upper bound on the mixing time, and the proposed approach improves the learning efficiency by keeping the mixing time of the Markov chains short. In particular, we propose a method of temporal-difference learning for estimating the gradient of the hitting time. Numerical experiments show that the proposed method outperforms conventional methods of PGRL.

In machine learning and statistics, probabilistic inference involving multimodal distributions is quite difficult. This is especially true in high dimensional problems, where most existing algorithms cannot easily move from one mode to another. To address this issue, we propose a novel Bayesian inference approach based on Markov Chain Monte Carlo. Our method can effectively sample from multimodal distributions, especially when the dimension is high and the modes are isolated. To this end, it exploits and modifies the Riemannian geometric properties of the target distribution to create \emph{wormholes} connecting modes in order to facilitate moving between them. Further, our proposed method uses the regeneration technique in order to adapt the algorithm by identifying new modes and updating the network of wormholes without affecting the stationary distribution. To find new modes, as opposed to rediscovering those previously identified, we employ a novel mode searching algorithm that explores a \emph{residual energy} function obtained by subtracting an approximate Gaussian mixture density (based on previously discovered modes) from the target density function.

Recent advances of kernel methods have yielded a framework for representing probabilities using a reproducing kernel Hilbert space, called kernel embedding of distributions. In this paper, we propose a Monte Carlo filtering algorithm based on kernel embeddings. The proposed method is applied to state-space models where sampling from the transition model is possible, while the observation model is to be learned from training samples without assuming a parametric model. As a theoretical basis of the proposed method, we prove consistency of the Monte Carlo method combined with kernel embeddings. Experimental results on synthetic models and real vision-based robot localization confirm the effectiveness of the proposed approach.

Nowadays, most recommender systems (RSs) mainly aim to suggest appropriate items for individuals. Due to the social nature of human beings, group activities have become an integral part of our daily life, thus motivating the study on group RS (GRS). However, most existing methods used by GRS make recommendations through aggregating individual ratings or individual predictive results rather than considering the collective features that govern user choices made within a group. As a result, such methods are heavily sensitive to data, hence they often fail to learn group preferences when the data are slightly inconsistent with predefined aggregation assumptions. To this end, we devise a novel GRS approach which accommodates both individual choices and group decisions in a joint model. More specifically, we propose a deep-architecture model built with collective deep belief networks and dual-wing restricted Boltzmann machines. With such a deep model, we can use high-level features, which are induced from lower-level features, to represent group preference so as to relieve the vulnerability of data. Finally, the experiments conducted on a real-world dataset prove the superiority of our deep model over other state-of-the-art methods.

Transfer learning considers related but distinct tasks defined on heterogenous domains and tries to transfer knowledge between these tasks to improve generalization performance. It is particularly useful when we do not have sufficient amount of labeled training data in some tasks, which may be very costly, laborious, or even infeasible to obtain. Instead, learning the tasks jointly enables us to effectively increase the amount of labeled training data. In this paper, we formulate a kernelized Bayesian transfer learning framework that is a principled combination of kernel-based dimensionality reduction models with task-specific projection matrices to find a shared subspace and a coupled classification model for all of the tasks in this subspace. Our two main contributions are: (i) two novel probabilistic models for binary and multiclass classification, and (ii) very efficient variational approximation procedures for these models. We illustrate the generalization performance of our algorithms on two different applications. In computer vision experiments, our method outperforms the state-of-the-art algorithms on nine out of 12 benchmark supervised domain adaptation experiments defined on two object recognition data sets. In cancer biology experiments, we use our algorithm to predict mutation status of important cancer genes from gene expression profiles using two distinct cancer populations, namely, patient-derived primary tumor data and in-vitro-derived cancer cell line data. We show that we can increase our generalization performance on primary tumors using cell lines as an auxiliary data source.

Maximization of submodular functions has wide applications in artificial intelligence and machine learning. In this paper, we propose a scalable learning algorithm for maximizing an adaptive submodular function. The key structural assumption in our solution is that the state of each item is distributed according to a generalized linear model, which is conditioned on the feature vector of the item. Our objective is to learn the parameters of this model. We analyze the performance of our algorithm, and show that its regret is polylogarithmic in time and linear in the number of features. Finally, we evaluate our solution on two problems, preference elicitation and adaptive face detection, and demonstrate that high-quality policies can be learned sample efficiently.

This paper presents an extension of the cascading Indian buffet process (CIBP) intended to learning arbitrary directed acyclic graph structures as opposed to the CIBP, which is limited to purely layered structures. The extended cascading Indian buffet process (eCIBP) essentially consists in adding an extra sampling step to the CIBP to generate connections between non-consecutive layers. In the context of graphical model structure learning, the proposed approach allows learning structures having an unbounded number of hidden random variables and automatically selecting the model complexity. We evaluated the extended process on multivariate density estimation and structure identification tasks by measuring the structure complexity and predictive performance. The results suggest the extension leads to extracting simpler graphs without scarifying predictive precision.