Annals of Mathematics and Artificial Intelligence最新文献

Contests are commonly used as a mechanism for eliciting effort and participation in multi-agent settings. Naturally, and much like with various other mechanisms, the information provided to the agents prior to and throughout the contest fundamentally influences its outcomes. In this paper we study the problem of information providing whenever the contest organizer does not initially hold the information and obtaining it is potentially costly. As the underlying contest mechanism for our model we use the binary contest, where contestants’ strategy is captured by their decision whether or not to participate in the contest in the first place. Here, it is often the case that the contest organizer can proactively obtain and provide contestants information related to their expected performance in the contest. We provide a comprehensive equilibrium analysis of the model, showing that even when such information is costless, it is not necessarily the case that the contest organizer will prefer to obtain and provide it to all agents, let alone when the information is costly.

The calibration problem in machine learning classification tasks arises when a model’s output score does not align with the ground truth observed probability of the target class. There exist several parametric and non-parametric post-processing methods that can help to calibrate an existing classifier. In this work, we focus on binary classification cases where the dataset is imbalanced, meaning that the negative target class significantly outnumbers the positive one. We propose new parametric calibration methods designed to this specific case and a new calibration measure focusing on the primary objective in imbalanced problems: detecting infrequent positive cases. Experiments on several datasets show that, for imbalanced problems, our approaches outperform state-of-the-art methods in many cases.

Conditional preference statements have been used to compactly represent preferences over combinatorial domains. They are at the core of CP-nets and their generalizations, and lexicographic preference trees. Several works have addressed the complexity of some queries (optimization, dominance in particular). We extend in this paper some of these results, and study other queries which have not been addressed so far, like equivalence, and transformations, like conditioning and variable elimination, thereby contributing to a knowledge compilation map for languages based on conditional preference statements. We also study the expressiveness and complexity of queries and transformations for generalized additive utilities.

The ecological psychologist James J. Gibson defined the notion of affordances to refer to what action possibilities environments offer to animals. In this paper, we show how (artificial) agents can discover and exploit affordances in a Multi-Agent System (MAS) environment to achieve their goals. To indicate to agents what affordances are present in their environment and whether it is likely that these may help the agents to achieve their objectives, the environment may expose signifiers while taking into account the current situation of the environment and of the agent. On this basis, we define a Signifier Exposure Mechanism that is used by the environment to compute which signifiers should be exposed to agents in order to permit agents to only perceive information about affordances that are likely to be relevant to them, and thereby increase their interaction efficiency. If this is successful, agents can interact with partially observable environments more efficiently because the signifiers indicate the affordances they can exploit towards given purposes. Signifiers thereby facilitate the exploration and the exploitation of MAS environments. Implementations of signifiers and of the Signifier Exposure Mechanism are presented within the context of a Hypermedia Multi-Agent System, and the utility of this approach is presented through the development of a scenario.

This paper describes experiments showing that some tasks in natural language processing (NLP) can already be performed using quantum computers, though so far only with small datasets. We demonstrate various approaches to topic classification. The first uses an explicit word-based approach, in which word-topic weights are implemented as fractional rotations of individual qubits, and a phrase is classified based on the accumulation of these weights onto a scoring qubit, using entangling quantum gates. This is compared with more scalable quantum encodings of word embedding vectors, which are used to compute kernel values in a quantum support vector machine: this approach achieved an average of 62% accuracy on classification tasks involving over 10000 words, which is the largest such quantum computing experiment to date. We describe a quantum probability approach to bigram modeling that can be applied to understand sequences of words and formal concepts, investigate a generative approximation to these distributions using a quantum circuit Born machine, and introduce an approach to ambiguity resolution in verb-noun composition using single-qubit rotations for simple nouns and 2-qubit entangling gates for simple verbs. The smaller systems presented have been run successfully on physical quantum computers, and the larger ones have been simulated. We show that statistically meaningful results can be obtained, but the quality of individual results varies much more using real datasets than using artificial language examples from previous quantum NLP research. Related NLP research is compared, partly with respect to contemporary challenges including informal language, fluency, and truthfulness.

Recent work has shown that Neural Ordinary Differential Equations (ODEs) can serve as generative models of images using the perspective of Continuous Normalizing Flows (CNFs). Such models offer exact likelihood calculation, and invertible generation/density estimation. In this work we introduce a Multi-Resolution variant of such models (MRCNF), by characterizing the conditional distribution over the additional information required to generate a fine image that is consistent with the coarse image. We introduce a transformation between resolutions that allows for no change in the log likelihood. We show that this approach yields comparable likelihood values for various image datasets, with improved performance at higher resolutions, with fewer parameters, using only one GPU. Further, we examine the out-of-distribution properties of MRCNFs, and find that they are similar to those of other likelihood-based generative models.

This paper develops a theory of clustering and coding that combines a geometric model with a probabilistic model in a principled way. The geometric model is a Riemannian manifold with a Riemannian metric, ({g}_{ij}(textbf{x})), which we interpret as a measure of dissimilarity. The probabilistic model consists of a stochastic process with an invariant probability measure that matches the density of the sample input data. The link between the two models is a potential function, (U(textbf{x})), and its gradient, (nabla U(textbf{x})). We use the gradient to define the dissimilarity metric, which guarantees that our measure of dissimilarity will depend on the probability measure. Finally, we use the dissimilarity metric to define a coordinate system on the embedded Riemannian manifold, which gives us a low-dimensional encoding of our original data.