| Total: 19
The paper investigates navigability with imperfect information. It shows that the properties of navigability with perfect recall are exactly those captured by Armstrong's axioms from database theory. If the assumption of perfect recall is omitted, then Armstrong's transitivity axiom is not valid, but it can be replaced by a weaker principle. The main technical results are soundness and completeness theorems for the logical systems describing properties of navigability with and without perfect recall.
We propose a novel variant of the UCB algorithm (referred to as Efficient-UCB-Variance (EUCBV)) for minimizing cumulative regret in the stochastic multi-armed bandit (MAB) setting. EUCBV incorporates the arm elimination strategy proposed in UCB-Improved, while taking into account the variance estimates to compute the arms' confidence bounds, similar to UCBV. Through a theoretical analysis we establish that EUCBV incurs a gap-dependent regret bound which is an improvement over that of existing state-of-the-art UCB algorithms (such as UCB1, UCB-Improved, UCBV, MOSS). Further, EUCBV incurs a gap-independent regret bound which is an improvement over that of UCB1, UCBV and UCB-Improved, while being comparable with that of MOSS and OCUCB. Through an extensive numerical study we show that EUCBV significantly outperforms the popular UCB variants (like MOSS, OCUCB, etc.) as well as Thompson sampling and Bayes-UCB algorithms.
Statistical relational AI (StarAI) aims at reasoning and learning in noisy domains described in terms of objects and relationships by combining probability with first-order logic. With huge advances in deep learning in the current years, combining deep networks with first-order logic has been the focus of several recent studies. Many of the existing attempts, however, only focus on relations and ignore object properties. The attempts that do consider object properties are limited in terms of modelling power or scalability. In this paper, we develop relational neural networks (RelNNs) by adding hidden layers to relational logistic regression (the relational counterpart of logistic regression). We learn latent properties for objects both directly and through general rules. Back-propagation is used for training these models. A modular, layer-wise architecture facilitates utilizing the techniques developed within deep learning community to our architecture. Initial experiments on eight tasks over three real-world datasets show that RelNNs are promising models for relational learning.
Consider a policymaker who wants to decide which intervention to perform in order to change a currently undesirable situation. The policymaker has at her disposal a team of experts, each with their own understanding of the causal dependencies between different factors contributing to the outcome. The policymaker has varying degrees of confidence in the experts’ opinions. She wants to combine their opinions in order to decide on the most effective intervention. We formally define the notion of an effective intervention, and then consider how experts’ causal judgments can be combined in order to determine the most effective intervention. We define a notion of two causal models being compatible, and show how compatible causal models can be combined. We then use it as the basis for combining experts causal judgments. We illustrate our approach on a number of real-life examples.
In the propositional setting, the marginal problem is to find a (maximum-entropy) distribution that has some given marginals. We study this problem in a relational setting and make the following contributions. First, we compare two different notions of relational marginals. Second, we show a duality between the resulting relational marginal problems and the maximum likelihood estimation of the parameters of relational models, which generalizes a well-known duality from the propositional setting. Third, by exploiting the relational marginal formulation, we present a statistically sound method to learn the parameters of relational models that will be applied in settings where the number of constants differs between the training and test data. Furthermore, based on a relational generalization of marginal polytopes, we characterize cases where the standard estimators based on feature's number of true groundings needs to be adjusted and we quantitatively characterize the consequences of these adjustments. Fourth, we prove bounds on expected errors of the estimated parameters, which allows us to lower-bound, among other things, the effective sample size of relational training data.
In planning algorithms and in other domains, there is often a need to run long computations that involve summations, maximizations and other operations on random variables, and to store intermediate results. In this paper, as a main motivating example, we elaborate on the case of estimating probabilities of meeting deadlines in hierarchical plans. A source of computational complexity, often neglected in the analysis of such algorithms, is that the support of the variables needed as intermediate results may grow exponentially along the computation. Therefore, to avoid exponential memory and time complexities, we need to trim these variables. This is similar, in a sense, to rounding intermediate results in numerical computations. Of course, to maintain the quality of algorithms, the trimming procedure should be efficient and it must maintain accuracy as much as possible. In this paper, we propose an optimal trimming algorithm with polynomial time and memory complexities for the purpose of estimating probabilities of deadlines in plans. More specifically, we show that our algorithm, given the needed size of the representation of the variable, provides the best possible approximation, where approximation accuracy is considered with a measure that fits the goal of estimating deadline meeting probabilities.
Robust reinforcement learning aims to produce policies that have strong guarantees even in the face of environments/transition models whose parameters have strong uncertainty. Existing work uses value-based methods and the usual primitive action setting. In this paper, we propose robust methods for learning temporally abstract actions, in the framework of options. We present a Robust Options Policy Iteration (ROPI) algorithm with convergence guarantees, which learns options that are robust to model uncertainty. We utilize ROPI to learn robust options with the Robust Options Deep Q Network (RO-DQN) that solves multiple tasks and mitigates model misspecification due to model uncertainty. We present experimental results which suggest that policy iteration with linear features may have an inherent form of robustness when using coarse feature representations. In addition, we present experimental results which demonstrate that robustness helps policy iteration implemented on top of deep neural networks to generalize over a much broader range of dynamics than non-robust policy iteration.
A multivariate Hawkes process is a class of marked point processes: A sample consists of a finite set of events of unbounded random size; each event has a real-valued time and a discrete-valued label (mark). It is self-excitatory: Each event causes an increase in the rate of other events (of either the same or a different label) in the (near) future. Prior work has developed methods for parameter estimation from complete samples. However, just as unobserved variables can increase the modeling power of other probabilistic models, allowing unobserved events can increase the modeling power of point processes. In this paper we develop a method to sample over the posterior distribution of unobserved events in a multivariate Hawkes process. We demonstrate the efficacy of our approach, and its utility in improving predictive power and identifying latent structure in real-world data.
Probabilistic topic models are popular unsupervised learning methods, including probabilistic latent semantic indexing (pLSI) and latent Dirichlet allocation (LDA). By now, their training is implemented on general purpose computers (GPCs), which are flexible in programming but energy-consuming. Towards low-energy implementations, this paper investigates their training on an emerging hardware technology called the neuromorphic multi-chip systems (NMSs). NMSs are very effective for a family of algorithms called spiking neural networks (SNNs). We present three SNNs to train topic models.The first SNN is a batch algorithm combining the conventional collapsed Gibbs sampling (CGS) algorithm and an inference SNN to train LDA. The other two SNNs are online algorithms targeting at both energy- and storage-limited environments. The two online algorithms are equivalent with training LDA by using maximum-a-posterior estimation and maximizing the semi-collapsed likelihood, respectively.They use novel, tailored ordinary differential equations for stochastic optimization. We simulate the new algorithms and show that they are comparable with the GPC algorithms, while being suitable for NMS implementation. We also propose an extension to train pLSI and a method to prune the network to obey the limited fan-in of some NMSs.
We consider sequential decision making problems under uncertainty, in which a user has a general idea of the task to achieve, and gives advice to an agent in charge of computing an optimal policy. Many different notions of advice have been proposed in somewhat different settings, especially in the field of inverse reinforcement learning and for resolution of Markov Decision Problems with Imprecise Rewards. Two key questions are whether the advice required by a specific method is natural for the user to give, and how much advice is needed for the agent to compute a good policy, as evaluated by the user. We give a unified view of a number of proposals made in the literature, and propose a new notion of advice, which corresponds to a user telling why she would take a given action in a given state. For all these notions, we discuss their naturalness for a user and the integration of advice. We then report on an experimental study of the amount of advice needed for the agent to compute a good policy. Our study shows in particular that continual interaction between the user and the agent is worthwhile, and sheds light on the pros and cons of each type of advice.
Probabilistic Sentential Decision Diagrams (PSDDs) have been proposed for learning tractable probability distributions from a combination of data and background knowledge (in the form of Boolean constraints). In this paper, we propose a variant on PSDDs, called conditional PSDDs, for representing a family of distributions that are conditioned on the same set of variables. Conditional PSDDs can also be learned from a combination of data and (modular) background knowledge. We use conditional PSDDs to define a more structured version of Bayesian networks, in which nodes can have an exponential number of states, hence expanding the scope of domains where Bayesian networks can be applied. Compared to classical PSDDs, the new representation exploits the independencies captured by a Bayesian network to decompose the learning process into localized learning tasks, which enables the learning of better models while using less computation. We illustrate the promise of conditional PSDDs and structured Bayesian networks empirically, and by providing a case study to the modeling of distributions over routes on a map.
Weight learning is a challenging problem in Markov Logic Networks (MLNs) due to the large size of the ground propositional probabilistic graphical model that underlies the first-order representation of MLNs. Though more sophisticated weight learning methods that use lifted inference have been proposed, such methods can typically scale up only in the absence of evidence, namely in generative weight learning. In discriminative learning, where the evidence typically destroys symmetries, existing approaches are lacking in scalability. In this paper, we propose a novel, intuitive approach for learning MLNs discriminatively by utilizing approximate symmetries. Specifically, we reduce the size of the training database by clustering approximately symmetric atoms together and selecting a representative atom from each cluster. However, each choice made from the clusters induces a different distribution, increasing the uncertainty in our learned model. To reduce this uncertainty, we learn a finite mixture model by stacking the different distributions, where the parameters of the model are learned using an EM approach. Our results on several benchmarks show that our approach is much more scalable and accurate as compared to existing state-of-the-art MLN learning methods.
The conditional value at risk (CVaR) is a popular risk measure which enables risk-averse decision making under uncertainty. We consider maximizing the CVaR of a continuous submodular function, an extension of submodular set functions to a continuous domain. One example application is allocating a continuous amount of energy to each sensor in a network, with the goal of detecting intrusion or contamination. Previous work allows maximization of the CVaR of a linear or concave function. Continuous submodularity represents a natural set of nonconcave functions with diminishing returns, to which existing techniques do not apply. We give a (1 - 1/e)-approximation algorithm for maximizing the CVaR of a monotone continuous submodular function. This also yields an algorithm for submodular set functions which produces a distribution over feasible sets with guaranteed CVaR. Experimental results in two sensor placement domains confirm that our algorithm substantially outperforms competitive baselines.
We introduce a theoretical model of information acquisition under resource limitations in a noisy environment. An agent must guess the truth value of a given Boolean formula φ after performing a bounded number of noisy tests of the truth values of variables in the formula. We observe that, in general, the problem of finding an optimal testing strategy for φ is hard, but we suggest a useful heuristic. The techniques we use also give insight into two apparently unrelated, but well-studied problems: (1) rational inattention (the optimal strategy may involve hardly ever testing variables that are clearly relevant to φ) and (2) what makes a formula hard to learn/remember.
Marginal MAP is a key task in Bayesian inference and decision-making. It is known to be very difficult in general, particularly because the evaluation of each MAP assignment requires solving an internal summation problem. In this paper, we propose a best-first search algorithm that provides anytime upper bounds for marginal MAP in graphical models. It folds the computation of external maximization and internal summation into an AND/OR tree search framework, and solves them simultaneously using a unified best-first search algorithm. The algorithm avoids some unnecessary computation of summation sub-problems associated with MAP assignments, and thus yields significant time savings. Furthermore, our algorithm is able to operate within limited memory. Empirical evaluation on three challenging benchmarks demonstrates that our unified best-first search algorithm using pre-compiled variational heuristics often provides tighter anytime upper bounds compared to those state-of-the-art baselines.
In this paper, we show that the recent integration of statistical models with deep recurrent neural networks provides a new way of formulating volatility (the degree of variation of time series) models that have been widely used in time series analysis and prediction in finance. The model comprises a pair of complementary stochastic recurrent neural networks: the generative network models the joint distribution of the stochastic volatility process; the inference network approximates the conditional distribution of the latent variables given the observables. Our focus here is on the formulation of temporal dynamics of volatility over time under a stochastic recurrent neural network framework. Experiments on real-world stock price datasets demonstrate that the proposed model generates a better volatility estimation and prediction that outperforms mainstream methods, e.g., deterministic models such as GARCH and its variants, and stochastic models namely the MCMC-based stochvol as well as the Gaussian-process-based, on average negative log-likelihood.
Selection and confounding biases are the two most common impediments to the applicability of causal inference methods in large-scale settings. We generalize the notion of backdoor adjustment to account for both biases and leverage external data that may be available without selection bias (e.g., data from census). We introduce the notion of adjustment pair and present complete graphical conditions for identifying causal effects by adjustment. We further design an algorithm for listing all admissible adjustment pairs in polynomial delay, which is useful for researchers interested in evaluating certain properties of some admissible pairs but not all (common properties include cost, variance, and feasibility to measure). Finally, we describe a statistical estimation procedure that can be performed once a set is known to be admissible, which entails different challenges in terms of finite samples.
Many real-world problems, such as Markov Logic Networks (MLNs) with evidence, can be represented as a highly symmetric graphical model perturbed by additional potentials. In these models, variational inference approaches that exploit exact model symmetries are often forced to ground the entire problem, while methods that exploit approximate symmetries (such as by constructing an over-symmetric approximate model) offer no guarantees on solution quality. In this paper, we present a method based on a lifted variant of the generalized dual decomposition (GenDD) for marginal MAP inference which provides a principled way to exploit symmetric sub-structures in a graphical model. We develop a coarse-to-fine inference procedure that provides any-time upper bounds on the objective. The upper bound property of GenDD provides a principled way to guide the refinement process, providing good any-time performance and eventually arriving at the ground optimal solution.
Rademacher complexity is often used to characterize the learnability of a hypothesis class and is known to be related to the class size. We leverage this observation and introduce a new technique for estimating the size of an arbitrary weighted set, defined as the sum of weights of all elements in the set. Our technique provides upper and lower bounds on a novel generalization of Rademacher complexity to the weighted setting in terms of the weighted set size. This generalizes Massart’s Lemma, a known upper bound on the Rademacher complexity in terms of the unweighted set size. We show that the weighted Rademacher complexity can be estimated by solving a randomly perturbed optimization problem, allowing us to derive high probability bounds on the size of any weighted set. We apply our method to the problems of calculating the partition function of an Ising model and computing propositional model counts (#SAT). Our experiments demonstrate that we can produce tighter bounds than competing methods in both the weighted and unweighted settings.