This paper considers the problem of cooperation between self-interested agents in acquiring better information regarding the nature of the different options and opportunities available to them. By sharing individual findings with others, the agents can potentially achieve a substantial improvement in overall and individual expected benefits. Unfortunately, it is well known that with self-interested agents equilibrium considerations often dictate solutions that are far from the fully cooperative ones, hence the agents do not manage to fully exploit the potential benefits encapsulated in such cooperation. In this paper we introduce, analyze and demonstrate the benefit of five methods aiming to improve cooperative information gathering. Common to all five that they constrain and limit the information sharing process. Nevertheless, the decrease in benefit due to the limited sharing is outweighed by the resulting substantial improvement in the equilibrium individual information gathering strategies. The equilibrium analysis given in the paper, which, in itself is an important contribution to the study of cooperation between self-interested agents, enables demonstrating that for a wide range of settings an improved individual expected benefit is achieved for all agents when applying each of the five methods.

What is the difference of a prediction that is made with a causal model and a non-causal model? Suppose we intervene on the predictor variables or change the whole environment. The predictions from a causal model will in general work as well under interventions as for observational data. In contrast, predictions from a non-causal model can potentially be very wrong if we actively intervene on variables. Here, we propose to exploit this invariance of a prediction under a causal model for causal inference: given different experimental settings (for example various interventions) we collect all models that do show invariance in their predictive accuracy across settings and interventions. The causal model will be a member of this set of models with high probability. This approach yields valid confidence intervals for the causal relationships in quite general scenarios. We examine the example of structural equation models in more detail and provide sufficient assumptions under which the set of causal predictors becomes identifiable. We further investigate robustness properties of our approach under model misspecification and discuss possible extensions. The empirical properties are studied for various data sets, including large-scale gene perturbation experiments.

This paper addresses the problem of active learning of a multi-output Gaussian process (MOGP) model representing multiple types of coexisting correlated environmental phenomena. In contrast to existing works, our active learning problem involves selecting not just the most informative sampling locations to be observed but also the types of measurements at each selected location for minimizing the predictive uncertainty (i.e., posterior joint entropy) of a target phenomenon of interest given a sampling budget. Unfortunately, such an entropy criterion scales poorly in the numbers of candidate sampling locations and selected observations when optimized. To resolve this issue, we first exploit a structure common to sparse MOGP models for deriving a novel active learning criterion. Then, we exploit a relaxed form of submodularity property of our new criterion for devising a polynomial-time approximation algorithm that guarantees a constant-factor approximation of that achieved by the optimal set of selected observations. Empirical evaluation on real-world datasets shows that our proposed approach outperforms existing algorithms for active learning of MOGP and single-output GP models.

Clustering is a common technique for statistical data analysis, widely used in data mining, machine learning, pattern recognition, image analysis, bioinformatics and cyber security. Conventional ("flat", "hard") clustering methods accept a finite metric space (O, d) as input and return a partition of O as their output. Hierarchical clustering (HC) methods have a different philosophy: their output is an entire hierarchy of partitions, called a dendrogram, capable of exhibiting multi-scale structure in the data set [1, 2]. Rather than fixing the required number of clusters in advance, as is common for many flat clustering algorithms, it is more informative to furnish a hierarchy of clusters, providing an opportunity to choose a partition at a scale most natural for the context of the task at hand. Many HC methods require linkage functions to provide a measure of dissimilarity between clusters (see [3, 4] for a fairly recent review). Some commonly used linkage functions are single linkage, complete linkage, average linkage, etc. The SLHC method, though suffering from the so called "chaining effect", remains popular for large scale applications [5] because of the low complexity of implementing it using minimum spanning trees (MST) [6].

As the name of the course suggests, this lecture note introduces readers to a neural network based approach to natural language understanding/processing. In order to make it as self-contained as possible, I spend much time on describing basics of machine learning and neural networks, only after which how they are used for natural languages is introduced. On the language front, I almost solely focus on language modelling and machine translation, two of which I personally find most fascinating and most fundamental to natural language understanding. After about a month of lectures and about 40 pages of writing this lecture note, I found this fascinating note [47] by Yoav Goldberg on neural network models for natural language processing. This note deals with wider topics on natural language processing with distributed representations in more details, and I highly recommend you to read it (hopefully along with this lecture note.)

To identify the location of objects of a particular class, a passive computer vision system generally processes all the regions in an image to finally output few regions. However, we can use structure in the scene to search for objects without processing the entire image. We propose a search technique that sequentially processes image regions such that the regions that are more likely to correspond to the query class object are explored earlier. We frame the problem as a Markov decision process and use an imitation learning algorithm to learn a search strategy. Since structure in the scene is essential for search, we work with indoor scene images as they contain both unary scene context information and object-object context in the scene. We perform experiments on the NYU-depth v2 dataset and show that the unary scene context features alone can achieve a significantly high average precision while processing only 20-25\% of the regions for classes like bed and sofa. By considering object-object context along with the scene context features, the performance is further improved for classes like counter, lamp, pillow and sofa.

Expectation maximization (EM) has recently been shown to be an efficient algorithm for learning finite-state controllers (FSCs) in large decentralized POMDPs (Dec-POMDPs). However, current methods use fixed-size FSCs and often converge to maxima that are far from optimal. This paper considers a variable-size FSC to represent the local policy of each agent. These variable-size FSCs are constructed using a stick-breaking prior, leading to a new framework called \emph{decentralized stick-breaking policy representation} (Dec-SBPR). This approach learns the controller parameters with a variational Bayesian algorithm without having to assume that the Dec-POMDP model is available. The performance of Dec-SBPR is demonstrated on several benchmark problems, showing that the algorithm scales to large problems while outperforming other state-of-the-art methods.

Finding high quality plans for large planning problems is hard. Although some current anytime planners are often able to improve plans quickly, they tend to reach a limit at which the plans produced are still very far from the best possible, but these planners fail to find any further improvement, even when given several hours of runtime. We present an approach to continuing plan quality optimisation at larger time scales, and its implementation in a system called BDPO2. Key to this approach is a decomposition into subproblems of improving parts of the current best plan. The decomposition is based on block deordering, a form of plan deordering which identifies hierarchical plan structure. BDPO2 can be seen as an application of the large neighbourhood search (LNS) local search strategy to planning, where the neighbourhood of a plan is defined by replacing one or more subplans with improved subplans. On-line learning is also used to adapt the strategy for selecting subplans and subplanners over the course of plan optimisation. Even starting from the best plans found by other means, BDPO2 is able to continue improving plan quality, often producing better plans than other anytime planners when all are given enough runtime. The best results, however, are achieved by a combination of different techniques working together.

Deep learning takes advantage of large datasets and computationally efficient training algorithms to outperform other approaches at various machine learning tasks. However, imperfections in the training phase of deep neural networks make them vulnerable to adversarial samples: inputs crafted by adversaries with the intent of causing deep neural networks to misclassify. In this work, we formalize the space of adversaries against deep neural networks (DNNs) and introduce a novel class of algorithms to craft adversarial samples based on a precise understanding of the mapping between inputs and outputs of DNNs. In an application to computer vision, we show that our algorithms can reliably produce samples correctly classified by human subjects but misclassified in specific targets by a DNN with a 97% adversarial success rate while only modifying on average 4.02% of the input features per sample. We then evaluate the vulnerability of different sample classes to adversarial perturbations by defining a hardness measure. Finally, we describe preliminary work outlining defenses against adversarial samples by defining a predictive measure of distance between a benign input and a target classification.

Gaussian process is a theoretically appealing model for nonparametric analysis, but its computational cumbersomeness hinders its use in large scale and the existing reduced-rank solutions are usually heuristic. In this work, we propose a novel construction of Gaussian process as a projection from fixed discrete frequencies to any continuous location. This leads to a valid stochastic process that has a theoretic support with the reduced rank in the spectral density, as well as a high-speed computing algorithm. Our method provides accurate estimates for the covariance parameters and concise form of predictive distribution for spatial prediction. For non-stationary data, we adopt the mixture framework with a customized spectral dependency structure. This enables clustering based on local stationarity, while maintains the joint Gaussianness. Our work is directly applicable in solving some of the challenges in the spatial data, such as large scale computation, anisotropic covariance, spatio-temporal modeling, etc. We illustrate the uses of the model via simulations and an application on a massive dataset.