We present ASP Modulo `Space-Time', a declarative representational and computational framework to perform commonsense reasoning about regions with both spatial and temporal components. Supported are capabilities for mixed qualitative-quantitative reasoning, consistency checking, and inferring compositions of space-time relations; these capabilities combine and synergise for applications in a range of AI application areas where the processing and interpretation of spatio-temporal data is crucial. The framework and resulting system is the only general KR-based method for declaratively reasoning about the dynamics of `space-time' regions as first-class objects. We present an empirical evaluation (with scalability and robustness results), and include diverse application examples involving interpretation and control tasks.

User Simulators are one of the major tools that enable offline training of task-oriented dialogue systems. For this task the Agenda-Based User Simulator (ABUS) is often used. The ABUS is based on hand-crafted rules and its output is in semantic form. Issues arise from both properties such as limited diversity and the inability to interface a text-level belief tracker. This paper introduces the Neural User Simulator (NUS) whose behaviour is learned from a corpus and which generates natural language, hence needing a less labelled dataset than simulators generating a semantic output. In comparison to much of the past work on this topic, which evaluates user simulators on corpus-based metrics, we use the NUS to train the policy of a reinforcement learning based Spoken Dialogue System. The NUS is compared to the ABUS by evaluating the policies that were trained using the simulators. Cross-model evaluation is performed i.e. training on one simulator and testing on the other. Furthermore, the trained policies are tested on real users. In both evaluation tasks the NUS outperformed the ABUS.

We propose a method to combine the interpretability and expressive power of firstorder logic with the effectiveness of neural network learning. In particular, we introduce a lifted framework in which first-order rules are used to describe the structure of a given problem setting. These rules are then used as a template for constructing a number of neural networks, one for each training and testing example. As the different networks corresponding to different examples share their weights, these weights can be efficiently learned using stochastic gradient descent. Our framework provides a flexible way for implementing and combining a wide variety of modelling constructs. In particular, the use of first-order logic allows for a declarative specification of latent relational structures, which can then be efficiently discovered in a given data set using neural network learning. Experiments on 78 relational learning benchmarks clearly demonstrate the effectiveness of the framework.

MAMMA-DSCS (mammary carcinoma decision support core system) is a prototype implementation designed for the support of decision processes for breast cancer (mammary carcinoma) treatment plans: Given a set of patient values, the system suggests different applicable treatment plans. As additional knowledge sources, MAMMA-DSCS uses external ontologies containing further information and correlations which are not directly tied to the tumor itself (e.g. toxicities, drug interactions). As a consequence, general knowledge like the substance composition of a specific therapy and its pharmacological hierarchy, can be separated from the knowledge about the applicability of therapies for a patient. The latter is encoded in an ASP program that is able to access the external ontologies and to take the obtained information into account for determining the set of all therapy plans that may be applied in a given situation. The ASP program models medical knowledge combining general guidelines and up-to-date expert knowledge for treating breast cancer on a very finegrained level, originating from a hospital in Germany.

In the present paper we describe the technology for translating algorithmic descriptions of discrete functions to SAT. The proposed methods and algorithms of translation are aimed at application to the problems of SAT-based cryptanalysis. In the theoretical part of the paper we justify the main principles of general reduction to SAT for discrete functions from a class containing the majority of functions employed in cryptography. Based on these principles we describe the Transalg software system, developed with SAT-based cryptanalysis specifics in mind. We show the results of applications of Transalg to construction of a number of attacks on various cryptographic functions. Some of the corresponding attacks are state of the art. In the paper we also present the vast experimental data, obtained using the SAT-solvers that took first places at the SAT-competitions in the recent several years.

We consider the use of no-regret algorithms to compute equilibria for particular classes of convex-concave games. While standard regret bounds would lead to convergence rates on the order of $O(T^{-1/2})$, recent work \citep{RS13,SALS15} has established $O(1/T)$ rates by taking advantage of a particular class of optimistic prediction algorithms. In this work we go further, showing that for a particular class of games one achieves a $O(1/T^2)$ rate, and we show how this applies to the Frank-Wolfe method and recovers a similar bound \citep{D15}. We also show that such no-regret techniques can even achieve a linear rate, $O(\exp(-T))$, for equilibrium computation under additional curvature assumptions.

We apply both distance-based (Jin and Matteson, 2017) and kernel-based (Pfister et al., 2016) mutual dependence measures to independent component analysis (ICA), and generalize dCovICA (Matteson and Tsay, 2017) to MDMICA, minimizing empirical dependence measures as an objective function in both deflation and parallel manners. Solving this minimization problem, we introduce Latin hypercube sampling (LHS) (McKay et al., 2000), and a global optimization method, Bayesian optimization (BO) (Mockus, 1994) to improve the initialization of the Newton-type local optimization method. The performance of MDMICA is evaluated in various simulation studies and an image data example. When the ICA model is correct, MDMICA achieves competitive results compared to existing approaches. When the ICA model is misspecified, the estimated independent components are less mutually dependent than the observed components using MDMICA, while they are prone to be even more mutually dependent than the observed components using other approaches.

We use an algebraic viewpoint, namely a matrix framework to deal with the problem of resource allocation under uncertainty in the context of a qualitative approach. Our basic qualitative data are a plausibility relation over the resources, a hierarchical relation over the agents and of course the preference that the agents have over the resources. With this data we propose a qualitative binary relation $\unrhd$ between allocations such that $\mathcal{F}\unrhd \mathcal{G}$ has the following intended meaning: the allocation $\mathcal{F}$ produces more or equal social welfare than the allocation $\mathcal{G}$. We prove that there is a family of allocations which are maximal with respect to $\unrhd$. We prove also that there is a notion of simple deal such that optimal allocations can be reached by sequences of simple deals. Finally, we introduce some mechanism for discriminating {optimal} allocations.

Generative Adversarial Nets (GANs) and Variational Auto-Encoders (VAEs) provide impressive image generations from Gaussian white noise, but the underlying mathematics are not well understood. We compute deep convolutional network generators by inverting a fixed embedding operator. Therefore, they do not require to be optimized with a discriminator or an encoder. The embedding is Lipschitz continuous to deformations so that generators transform linear interpolations between input white noise vectors into deformations between output images. This embedding is computed with a wavelet Scattering transform. Numerical experiments demonstrate that the resulting Scattering generators have similar properties as GANs or VAEs, without learning a discriminative network or an encoder.

Recent work has managed to learn cross-lingual word embeddings without parallel data by mapping monolingual embeddings to a shared space through adversarial training. However, their evaluation has focused on favorable conditions, using comparable corpora or closely-related languages, and we show that they often fail in more realistic scenarios. This work proposes an alternative approach based on a fully unsupervised initialization that explicitly exploits the structural similarity of the embeddings, and a robust self-learning algorithm that iteratively improves this solution. Our method succeeds in all tested scenarios and obtains the best published results in standard datasets, even surpassing previous supervised systems. Our implementation is released as an open source project at https://github.com/artetxem/vecmap