We propose a new Bayesian model for reliable aggregatio of crowdsourced estimates of real-valued quantities in participatory sensing applications. Existing approaches focus on probabilistic modelling of user’s reliability as the key to accurate aggregation. However, these are either limited to estimating discrete quantities, or require a significant number of reports from each user to accurately model their reliability. To mitigate these issues, we adopt a community-based approach, which reduces the data required to reliably aggregate real-valued estimates, by leveraging correlations between the reporting behaviour of users belonging to different communities. As a result, our method is up to 16.6% more accurate than existing state-of-the-art methods and is up to 49% more effective under data sparsity when used to estimate Wi-Fi hotspot locations in a real-world crowdsourcing application.

The recently proposed Gaussian process dynamical models (GPDMs) have been successfully applied to time series modeling. There are four learning algorithms for GPDMs: maximizing a posterior (MAP), fixing the kernel hyperparameters α _ (Fix.α _ ), balanced GPDM (B-GPDM) and two-stage MAP (T.MAP), which are designed for model training with complete data. When data are incomplete, GPDMs reconstruct the missing data using a function of the latent variables before parameter updates, which, however, may cause cumulative errors. In this paper, we present four new algorithms (MAP+, Fix.α + , B-GPDM+ and T.MAP+) for learning GPDMs with incomplete training data and a new conditional model (CM+) for recovering incomplete test data. Our methods adopt the Bayesian framework and can fully and properly use the partially observed data. We conduct experiments on incomplete motion capture data (walk, run, swing and multiple-walker) and make comparisons with the existing four algorithms as well as k-NN, spline interpolation and VGPDS. Our methods perform much better on both training with incomplete data and recovering incomplete test data.

We present Saul, a new probabilistic programming language designed to address some of the shortcomings of programming languages that aim at advancing and simplifying the development of AI systems. Such languages need to interact with messy, naturally occurring data, to allow a programmer to specify what needs to be done at an appropriate level of abstraction rather than at the data level, to be developed on a solid theory that supports moving to and reasoning at this level of abstraction and, finally, to support flexible integration of these learning and inference models within an application program. Saul is an object-functional programming language written in Scala that facilitates these by (1) allowing a programmer to learn, name and manipulate named abstractions over relational data; (2) supporting seamless incorporation of trainable (probabilistic or discriminative) components into the program, and (3) providing a level of inference over trainable models to support composition and make decisions that respect domain and application constraints. Saul is developed over a declaratively defined relational data model, can use piecewise learned factor graphs with declaratively specified learning and inference objectives, and it supports inference over probabilistic models augmented with declarative knowledge-based constraints.We describe the key constructs of Saul and exemplify its use in developing applications that require relational feature engineering and structured output prediction.

The conventional model for online planning under uncertainty assumes that an agent can stop and plan without incurring costs for the time spent planning. However, planning time is not free in most real-world settings. For example, an autonomous drone is subject to nature's forces, like gravity, even while it thinks, and must either pay a price for counteracting these forces to stay in place, or grapple with the state change caused by acquiescing to them. Policy optimization in these settings requires metareasoning---a process that trades off the cost of planning and the potential policy improvement that can be achieved. We formalize and analyze the metareasoning problem for Markov Decision Processes (MDPs). Our work subsumes previously studied special cases of metareasoning and shows that in the general case, metareasoning is at most polynomially harder than solving MDPs with any given algorithm that disregards the cost of thinking. For reasons we discuss, optimal general metareasoning turns out to be impractical, motivating approximations. We present approximate metareasoning procedures which rely on special properties of the BRTDP planning algorithm and explore the effectiveness of our methods on a variety of problems.

In this work, we investigate the bidirectional mutual interactions (BMI) between users' activities and user-user relationships on social networking sites. We analyze and study the fundamental mechanism that drives the characteristics and dynamics of BMI is the underlying social influence. We make an attempt at a unified probabilistic approach, called joint activity and relation (JAR), for modeling and predicting users' activities and user-user relationships simultaneously in a single coherent framework. Instead of incorporating social influence in an ad hoc manner, we show that social influence can be captured quantitatively. Based on JAR, we learn social influence between users and users' personal preferences for both user activity prediction and user-user relation discovery through statistical inference. To address the challenges of the introduced multiple layers of hidden variables in JAR, we propose a new learning algorithm based on expectation maximization (EM) and we further propose a powerful and efficient generalization of the EM based algorithm for model fitting.We show that JAR exploits mutual interactions and benefits, by taking advantage of the learned social influence and users' personal preferences, for enhanced user activity prediction and user-user relation discovery. We further experiment with real world dataset to verify the claimed advantages achieving substantial performance gains.

The classical probabilistic entailment problem is to We apply the family of minimal violation measures from determine upper and lower bounds on the probability [Potyka, 2014] since they allow us to extend the classical notion of formulas, given a consistent set of probabilistic of models of a probabilistic knowledge base to inconsistent assertions. We generalize this problem ones. Intuitively, the generalized models are those probability by omitting the consistency assumption and, thus, functions that minimally violate the knowledge base provide a general framework for probabilistic reasoning [Potyka and Thimm, 2014]. We incorporate integrity constraints under inconsistency. To do so, we utilize and study a family of generalized entailment problems inconsistency measures to determine probability for probabilistic knowledge bases. More specifically, functions that are closest to satisfying the knowledge the contributions of this work are as follows: base. We illustrate our approach on several 1. We introduce the computational problem of generalized examples and show that it has both nice formal and entailment with integrity constraints in probabilistic logics computational properties.

In recommendation systems, probabilistic matrix factorization (PMF) is a state-of-the-art collaborative filtering method by determining the latent features to represent users and items. However, two major issues limiting the usefulness of PMF are the sparsity problem and long-tail distribution. Sparsity refers to the situation that the observed rating data are sparse, which results in that only part of latent features are informative for describing each item/user. Long tail distribution implies that a large fraction of items have few ratings. In this work, we propose a sparse probabilistic matrix factorization method (SPMF) by utilizing a Laplacian distribution to model the item/user factor vector. Laplacian distribution has ability to generate sparse coding, which is beneficial for SPMF to distinguish the relevant and irrelevant latent features with respect to each item/user. Meanwhile, the tails in Laplacian distribution are comparatively heavy, which is rewarding for SPMF to recommend the tail items. Furthermore, a distributed Gibbs sampling algorithm is developed to efficiently train the proposed sparse probabilistic model. A series of experiments on Netfilix and Movielens datasets have been conducted to demonstrate that SPMF outperforms the existing PMF and its extended version Bayesian PMF (BPMF), especially for the recommendation of tail items.

Large multilayer neural networks trained with backpropagation have recently achieved state-of-the-art results in a wide range of problems. However, using backprop for neural net learning still has some disadvantages, e.g., having to tune a large number of hyperparameters to the data, lack of calibrated probabilistic predictions, and a tendency to overfit the training data. In principle, the Bayesian approach to learning neural networks does not have these problems. However, existing Bayesian techniques lack scalability to large dataset and network sizes. In this work we present a novel scalable method for learning Bayesian neural networks, called probabilistic backpropagation (PBP). Similar to classical backpropagation, PBP works by computing a forward propagation of probabilities through the network and then doing a backward computation of gradients. A series of experiments on ten real-world datasets show that PBP is significantly faster than other techniques, while offering competitive predictive abilities. Our experiments also show that PBP provides accurate estimates of the posterior variance on the network weights.

Nicod's criterion states that observing a black raven is evidence for the hypothesis H that all ravens are black. We show that Solomonoff induction does not satisfy Nicod's criterion: there are time steps in which observing black ravens decreases the belief in H. Moreover, while observing any computable infinite string compatible with H, the belief in H decreases infinitely often when using the unnormalized Solomonoff prior, but only finitely often when using the normalized Solomonoff prior. We argue that the fault is not with Solomonoff induction; instead we should reject Nicod's criterion.

Gaussian processes (GP) are powerful tools for probabilistic modeling purposes. They can be used to define prior distributions over latent functions in hierarchical Bayesian models. The prior over functions is defined implicitly by the mean and covariance function, which determine the smoothness and variability of the function. The inference can then be conducted directly in the function space by evaluating or approximating the posterior process. Despite their attractive theoretical properties GPs provide practical challenges in their implementation. GPstuff is a versatile collection of computational tools for GP models compatible with Linux and Windows MATLAB and Octave. It includes, among others, various inference methods, sparse approximations and tools for model assessment. In this work, we review these tools and demonstrate the use of GPstuff in several models.