Is neural IR mostly hype? In a recent SIGIR Forum article, Lin expressedskepticism that neural ranking models were actually improving ad hoc retrievaleffectiveness in limited data scenarios. He... | Continue reading
Gossip protocols are programs used in a setting in which each agent holds asecret and the aim is to reach a situation in which all agents know allsecrets. Such protocols rely on a point-to-point... | Continue reading
We introduce a conceptually simple and effective method to quantify thesimilarity between relations in knowledge bases. Specifically, our approach isbased on the divergence between the... | Continue reading
We present a new point-based approach for modeling complex scenes. Theapproach uses a raw point cloud as the geometric representation of a scene, andaugments each point with a learnable neural... | Continue reading
Transportation service providers that dispatch drivers and vehicles to ridersstart to support both on-demand ride requests posted in real time and ridesscheduled in advance, leading to new... | Continue reading
Generative adversarial networks (GAN) are a powerful subclass of generativemodels. Despite a very rich research activity leading to numerous interestingGAN algorithms, it is still very hard to... | Continue reading
We introduce DiscEval, a compilation of $11$ evaluation datasets with a focuson discourse, that can be used for evaluation of English Natural LanguageUnderstanding when considering meaning as... | Continue reading
In order for robots and other artificial agents to efficiently learn toperform useful tasks defined by an end user, they must understand not only thegoals of those tasks, but also the structure... | Continue reading
Magic sets are a Datalog to Datalog rewriting technique to optimize queryanswering. The rewritten program focuses on a portion of the stable model(s) ofthe input program which is sufficient to... | Continue reading
Specifying a complete domain model is time-consuming, which has been abottleneck of AI planning technique application in many real-world scenarios.Most classical domain-model learning approaches... | Continue reading
This is a set of lecture notes suitable for a Master's course on quantumcomputation and information from the perspective of theoretical computerscience. The first version was written in 2011,... | Continue reading
In symbolic planning systems, the knowledge on the domain is commonlyprovided by an expert. Recently, an automatic abstraction procedure has beenproposed in the literature to create a Planning... | Continue reading
We present SentiMATE, a novel end-to-end Deep Learning model for Chess,employing Natural Language Processing that aims to learn an effectiveevaluation function assessing move quality. This... | Continue reading
Neural networks have become an increasingly popular tool for solving manyreal-world problems. They are a general framework for differentiableoptimization which includes many other machine... | Continue reading
Designing effective low-level robot controllers often entailplatform-specific implementations that require manual heuristic parametertuning, significant system knowledge, or long design times.... | Continue reading
In this report we review memory-based meta-learning as a tool for buildingsample-efficient strategies that learn from past experience to adapt to anytask within a target class. Our goal is to... | Continue reading
Many road accidents occur due to distracted drivers. Today, driver monitoringis essential even for the latest autonomous vehicles to alert distracteddrivers in order to take over control of the... | Continue reading
We consider cosmological evolution from the perspective of quantuminformation. We present a quantum circuit model for the expansion of a comovingregion of space, in which initially-unentangled... | Continue reading
The Gleason score is the most important prognostic marker for prostate cancerpatients but suffers from significant inter-observer variability. We developeda fully automated deep learning system... | Continue reading
Compared with single image based crowd counting, video provides thespatial-temporal information of the crowd that would help improve therobustness of crowd counting. But translation, rotation... | Continue reading
Object tracking has important application in assistive technologies forpersonalized monitoring. Recent trackers choosing AlexNet as their backbone toextract features have gained great success.... | Continue reading
We study the transition in the functional networks that characterize thehuman brains' conscious-state to an unconscious subliminal state of perceptionby using k-core percolation. We find that... | Continue reading
We develop an agent-based simulation of the catastrophe insurance andreinsurance industry and use it to study the problem of risk model homogeneity.The model simulates the balance sheets of... | Continue reading
A common trait of complex systems is that they can be represented by means ofa network of interacting parts. It is, in fact, the network organisation (morethan the parts) what largely conditions... | Continue reading
The powerful representation capacity of deep learning has made it inevitablefor the underwater image enhancement community to employ its potential. Theexploration of deep underwater image... | Continue reading
CNNs have made an undeniable impact on computer vision through the ability tolearn high-capacity models with large annotated training sets. One of theirremarkable properties is the ability to... | Continue reading
In this paper, we propose a regularized graph neural network (RGNN) forEEG-based emotion recognition. EEG signals measure the neuronal activities ondifferent brain regions via electrodes... | Continue reading
Meta-training has been empirically demonstrated to be the most effectivepre-training method for few-shot learning of medical image classifiers (i.e.,classifiers modeled with small training... | Continue reading
Nowadays, autonomous driving cars have become commercially available.However, the safety of a self-driving car is still a challenging problem thathas not been well studied. Motion prediction is... | Continue reading
Proof-of-Space provides an intriguing alternative for consensus protocol ofpermissionless blockchains due to its recyclable nature and the potential tosupport multiple chains simultaneously.... | Continue reading
Rhythm Dungeon is a rhythm game which leverages the blockchain as a sharedopen database. During the gaming session, the player explores a roguelikedungeon by inputting specific sequences in time... | Continue reading
Significant work is being done to develop the math and tools necessary tobuild provable defenses, or at least bounds, against adversarial attacks ofneural networks. In this work, we argue that... | Continue reading
The inherent difficulty of knowledge specification and the lack of trainedspecialists are some of the key obstacles on the way to making intelligentsystems based on the knowledge representation... | Continue reading
Smart reply systems have been developed for various messaging platforms. Inthis paper, we introduce Uber's smart reply system: one-click-chat (OCC), whichis a key enhanced feature on top of the... | Continue reading
In this work, we consider the medical concept normalization problem, i.e.,the problem of mapping a health-related entity mention in a free-form text to aconcept in a controlled vocabulary,... | Continue reading
This article is a linguistic investigation of a neural parser. We look attransitivity and agreement information of auxiliary verb constructions (AVCs)in comparison to finite main verbs (FMVs).... | Continue reading
In data dominated systems and applications, a concept of representing wordsin a numerical format has gained a lot of attention. There are a few approachesused to generate such a representation.... | Continue reading
An elastic-visco-plastic thermomechanical model for the simulation of coldforming and subsequent sintering of ceramic powders is introduced and based onmicromechanical modelling of the... | Continue reading
An effective way to maximize code coverage in software tests is throughdynamic symbolic execution$-$a technique that uses constraint solving tosystematically explore a program's state space. We... | Continue reading
Foundational verification allows programmers to build software which has beenempirically shown to have high levels of assurance in a variety of importantdomains. However, the cost of producing... | Continue reading
The evolution and development of events have their own basic principles,which make events happen sequentially. Therefore, the discovery of suchevolutionary patterns among events are of great... | Continue reading
For a robot to learn a good policy, it often requires expensive equipment(such as sophisticated sensors) and a prepared training environment conduciveto learning. However, it is seldom possible... | Continue reading
Comprehensibility of source code is strongly affected by identifier names,therefore software developers need to give good (e.g. meaningful but short)names to identifiers. On the other hand,... | Continue reading
We propose a new way of constructing invertible neural networks by combiningsimple building blocks with a novel set of composition rules. This leads to arich set of invertible architectures,... | Continue reading
Scientific computing is increasingly incorporating the advancements inmachine learning and the ability to work with large amounts of data. At thesame time, machine learning models are becoming... | Continue reading
Transformer is a popularly used neural network architecture, especially forlanguage understanding. We introduce an extended and unified architecture whichcan be used for tasks involving a... | Continue reading
Fundamental mathematical constants like $e$ and $π$ are ubiquitous indiverse fields of science, from abstract mathematics and geometry to physics,biology and chemistry. Nevertheless, for... | Continue reading
The widespread use of conversational and question answering systems made itnecessary to improve the performances of speaker intent detection andunderstanding of related semantic slots, i.e.,... | Continue reading