Self-organized Hierarchical Softmax

We propose a new self-organizing hierarchical softmax formulation for neural-network-based language models over large vocabularies. Instead of using a predefined hierarchical structure, our approach is capable of learning word clusters with clear syntactical and semantic meaning during the language model training process. We provide experiments on standard benchmarks for language modeling and sentence compression tasks. We find that this approach is as fast as other efficient softmax approximations, while achieving comparable or even better performance relative to similar full softmax models.


Guiding Reinforcement Learning Exploration Using Natural Language

In this work we present a technique to use natural language to help reinforcement learning generalize to unseen environments. This technique uses neural machine translation to learn associations between natural language behavior descriptions and state-action information. We then use this learned model to guide agent exploration to make it more effective at learning in unseen environments. We evaluate this technique using the popular arcade game, Frogger, under ideal and non-ideal conditions. This evaluation shows that our modified policy shaping algorithm improves over a Q-learning agent as well as a baseline version of policy shaping.


Change Point Detection with Optimal Transport and Geometric Discrepancy

We present novel retrospective change point detection approach based on optimal transport and geometric discrepancy. The method does not require any parametric assumptions about distributions separated by change points. It can be used both for single and multiple change point detection and estimation, while the number of change points is either known or unknown. This result is achieved by construction of a certain sliding window statistic from which change points can be derived with elementary convex geometry in a specific Hilbert space. The work is illustrated with computational examples, both artificially constructed and based on actual data.


Multi-Robot Transfer Learning: A Dynamical System Perspective

Multi-robot transfer learning allows a robot to use data generated by a second, similar robot to improve its own behavior. The potential advantages are reducing the time of training and the unavoidable risks that exist during the training phase. Transfer learning algorithms aim to find an optimal transfer map between different robots. In this paper, we investigate, through a theoretical study of single-input single-output (SISO) systems, the properties of such optimal transfer maps. We first show that the optimal transfer learning map is, in general, a dynamic system. The main contribution of the paper is to provide an algorithm for determining the properties of this optimal dynamic map including its order and regressors (i.e., the variables it depends on). The proposed algorithm does not require detailed knowledge of the robots’ dynamics, but relies on basic system properties easily obtainable through simple experimental tests. We validate the proposed algorithm experimentally through an example of transfer learning between two different quadrotor platforms. Experimental results show that an optimal dynamic map, with correct properties obtained from our proposed algorithm, achieves 60-70% reduction of transfer learning error compared to the cases when the data is directly transferred or transferred using an optimal static map.


A Knowledge-Based Analysis of the Blockchain Protocol

At the heart of the Bitcoin is a blockchain protocol, a protocol for achieving consensus on a public ledger that records bitcoin transactions. To the extent that a blockchain protocol is used for applications such as contract signing and making certain transactions (such as house sales) public, we need to understand what guarantees the protocol gives us in terms of agents’ knowledge. Here, we provide a complete characterization of agent’s knowledge when running a blockchain protocol using a variant of common knowledge that takes into account the fact that agents can enter and leave the system, it is not known which agents are in fact following the protocol (some agents may want to deviate if they can gain by doing so), and the fact that the guarantees provided by blockchain protocols are probabilistic. We then consider some scenarios involving contracts and show that this level of knowledge suffices for some scenarios, but not others.


Bayesian Decision Theory and Stochastic Independence

Stochastic independence has a complex status in probability theory. It is not part of the definition of a probability measure, but it is nonetheless an essential property for the mathematical development of this theory. Bayesian decision theorists such as Savage can be criticized for being silent about stochastic independence. From their current preference axioms, they can derive no more than the definitional properties of a probability measure. In a new framework of twofold uncertainty, we introduce preference axioms that entail not only these definitional properties, but also the stochastic independence of the two sources of uncertainty. This goes some way towards filling a curious lacuna in Bayesian decision theory.


On the average of probability distributions of a discrete Markov chain

Let X_n be a discrete time Markov chain with state space S and initial probability distribution \mu^{(0)} = (P(X_0=i_1),P(X_0=i_2),\cdots,). What is the probability of choosing in random some k \in \mathbb{N} with k \leq n such that X_k = j where j \in S? This probability is the average \frac{1}{n} \sum_{k=1}^n \mu^{(k)}_j where \mu^{(k)}_j = P(X_k = j). In this note we will study the limit of this average without assuming that the chain is irreducible. Finally, we study the limit of the average \frac{1}{n} \sum_{k=1}^n g(X_k) where g is a given function for a general Markov chain not necessarily irreducible.


STN-OCR: A single Neural Network for Text Detection and Text Recognition

Detecting and recognizing text in natural scene images is a challenging, yet not completely solved task. In recent years several new systems that try to solve at least one of the two sub-tasks (text detection and text recognition) have been proposed. In this paper we present STN-OCR, a step towards semi-supervised neural networks for scene text recognition, that can be optimized end-to-end. In contrast to most existing works that consist of multiple deep neural networks and several pre-processing steps we propose to use a single deep neural network that learns to detect and recognize text from natural images in a semi-supervised way. STN-OCR is a network that integrates and jointly learns a spatial transformer network, that can learn to detect text regions in an image, and a text recognition network that takes the identified text regions and recognizes their textual content. We investigate how our model behaves on a range of different tasks (detection and recognition of characters, and lines of text). Experimental results on public benchmark datasets show the ability of our model to handle a variety of different tasks, without substantial changes in its overall network structure.


Detecting and Explaining Causes From Text For a Time Series Event

Explaining underlying causes or effects about events is a challenging but valuable task. We define a novel problem of generating explanations of a time series event by (1) searching cause and effect relationships of the time series with textual data and (2) constructing a connecting chain between them to generate an explanation. To detect causal features from text, we propose a novel method based on the Granger causality of time series between features extracted from text such as N-grams, topics, sentiments, and their composition. The generation of the sequence of causal entities requires a commonsense causative knowledge base with efficient reasoning. To ensure good interpretability and appropriate lexical usage we combine symbolic and neural representations, using a neural reasoning algorithm trained on commonsense causal tuples to predict the next cause step. Our quantitative and human analysis show empirical evidence that our method successfully extracts meaningful causality relationships between time series with textual features and generates appropriate explanation between them.


Delegated causality of complex systems

The article introduces a simple but subtle, overlooked kind of causality provoked by critical dynamical systems with rich behavior and moderate sensitivity to the environment. It is argued that conspicuously complex natural systems build up on interactions of this provoked, delegated causality.


A Family of Metrics for Clustering Algorithms

We give the motivation for scoring clustering algorithms and a metric M : A \rightarrow \mathbb{N} from the set of clustering algorithms to the natural numbers which we realize as \begin{equation} M(A) = \sum_i \alpha_i |f_i – \beta_i|^{w_i} \end{equation} where \alpha_i,\beta_i,w_i are parameters used for scoring the feature f_i, which is computed empirically.. We give a method by which one can score features such as stability, noise sensitivity, etc and derive the necessary parameters. We conclude by giving a sample set of scores.


Multi-Stakeholder Recommendation: Applications and Challenges

Recommender systems have been successfully applied to assist decision making by producing a list of item recommendations tailored to user preferences. Traditional recommender systems only focus on optimizing the utility of the end users who are the receiver of the recommendations. By contrast, multi-stakeholder recommendation attempts to generate recommendations that satisfy the needs of both the end users and other parties or stakeholders. This paper provides an overview and discussion about the multi-stakeholder recommendations from the perspective of practical applications, available data sets, corresponding research challenges and potential solutions.


Robust Physical-World Attacks on Machine Learning Models

Deep neural network-based classifiers are known to be vulnerable to adversarial examples that can fool them into misclassifying their input through the addition of small-magnitude perturbations. However, recent studies have demonstrated that such adversarial examples are not very effective in the physical world–they either completely fail to cause misclassification or only work in restricted cases where a relatively complex image is perturbed and printed on paper. In this paper we propose a new attack algorithm–Robust Physical Perturbations (RP2)– that generates perturbations by taking images under different conditions into account. Our algorithm can create spatially-constrained perturbations that mimic vandalism or art to reduce the likelihood of detection by a casual observer. We show that adversarial examples generated by RP2 achieve high success rates under various conditions for real road sign recognition by using an evaluation methodology that captures physical world conditions. We physically realized and evaluated two attacks, one that causes a Stop sign to be misclassified as a Speed Limit sign in 100% of the testing conditions, and one that causes a Right Turn sign to be misclassified as either a Stop or Added Lane sign in 100% of the testing conditions.


Polarization-Division Multiplexing Based on the Nonlinear Fourier Transform
Cognitive Hierarchy and Voting Manipulation
Pileup Mitigation with Machine Learning (PUMML)
Enforcing Constraints on Outputs with Unconstrained Inference
Communication versus Computation: Duality for multiple access channels and source coding
Robust Rigid Point Registration based on Convolution of Adaptive Gaussian Mixture Models
Optimizing Filter Size in Convolutional Neural Networks for Facial Action Unit Recognition
Sharpening Jensen’s Inequality
Learning a Target Sample Re-Generator for Cross-Database Micro-Expression Recognition
On the complexity of the projective splitting and Spingarn’s methods for the sum of two maximal monotone operators
On the packing numbers in graphs
Temporal dynamics of semantic relations in word embeddings: an application to predicting armed conflict participants
Non-existence of partial difference sets of order 8p^3 in Abelian groups
A Tale of Two DRAGGNs: A Hybrid Approach for Interpreting Action-Oriented and Goal-Oriented Instructions
The net number of US House seats won by partisan gerrymandering
Context-aware Single-Shot Detector
A Naive Algorithm for Feedback Vertex Set
The distance Laplacian spectral radius of unicyclic graphs
Vertex Deletion Problems on Chordal Graphs
Adaptive and Resilient Revenue Maximizing Resource Allocation and Pricing in Cloud Computing Environments
Extended Comparisons of Best Subset Selection, Forward Stepwise Selection, and the Lasso
Anytime Exact Belief Propagation
A Jointly Learned Deep Architecture for Facial Attribute Analysis and Face Detection in the Wild
Novel reformulations and efficient algorithm for the generalized trust region subproblem
A minimaj-preserving crystal on ordered multiset partitions
Signal and Noise Statistics Oblivious Sparse Reconstruction using OMP/OLS
Determining Semantic Textual Similarity using Natural Deduction Proofs
Ultra-low-power Wireless Streaming Cameras
Analysis of Deformation Fields in Spatio-temporal CBCT images of lungs for radiotherapy patients
Exploiting Web Images for Weakly Supervised Object Detection
Algebraic Relations and Triangulation of Unlabeled Image Points
An Improved Subsumption Testing Algorithm for the Optimal-Size Sorting Network Problem
Learning Audio Sequence Representations for Acoustic Event Classification
A Quantum Approach to Subset-Sum and Similar Problems
Common Knowledge in a Logic of Gossips
A Logic for Global and Local Announcements
Relaxing Exclusive Control in Boolean Games
A New Game Equivalence and its Modal Logic
From Type Spaces to Probability Frames and Back, via Language
Rationalizability and Epistemic Priority Orderings
Preservation of Semantic Properties during the Aggregation of Abstract Argumentation Frameworks
Binary Voting with Delegable Proxy: An Analysis of Liquid Democracy
Games With Tolerant Players
What Drives People’s Choices in Turn-Taking Games, if not Game-Theoretic Rationality?
An Epistemic Foundation for Authentication Logics (Extended Abstract)
Group Recommendations: Axioms, Impossibilities, and Random Walks
Together We Know How to Achieve: An Epistemic Logic of Know-How (Extended Abstract)
Condorcet’s Principle and the Preference Reversal Paradox
Self-confirming Games: Unawareness, Discovery, and Equilibrium
Argument-based Belief in Topological Structures
Reconciling Bayesian Epistemology and Narration-based Approaches to Judiciary Fact-finding
A New Modal Framework for Epistemic Logic
Existence and continuity of the flow constant in first passage percolation
An Improved Epsilon Constraint-handling Method in MOEA/D for CMOPs with Large Infeasible Regions
Ramsey Spanning Trees and their Applications
Proceedings of Workshop AEW10: Concepts in Information Theory and Communications
Toplogical Data Analysis of Clostridioides difficile Infection and Fecal Microbiota Transplantation
An Evolutionary Stochastic-Local-Search Framework for One-Dimensional Cutting-Stock Problems
Analysis of Italian Word Embeddings
Integrability of Liouville theory: proof of the DOZZ Formula
A note on surjectivity of piecewise affine mappings
Bayesian inference for Stable Levy driven Stochastic Differential Equations with high-frequency data
On σ-LCD codes
A note on minimal dispersion of point sets in the unit cube
Impact of Correlation between Interferers on Coverage Probability and rate in Cellular Systems
Nearest Common Ancestors: Universal Trees and Improved Labeling Schemes
A Comparative Study of the Clinical use of Motion Analysis from Kinect Skeleton Data
Representation-Aggregation Networks for Segmentation of Multi-Gigapixel Histology Images
Morphisms of Butson classes
Food Ingredients Recognition through Multi-label Learning
Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewards
A note on strong approximation of SDEs with smooth coefficients that have at most linearly growing derivatives
A Downsampled Variant of ImageNet as an Alternative to the CIFAR datasets
Max K-armed bandit: On the ExtremeHunter algorithm and beyond
Serious Games Application for Memory Training Using Egocentric Images
A stochastic maximal inequality, strict countability, and infinite-dimensional martingales
Many-body localization for randomly interacting bosons
LCD codes over ${\mathbb F}_q $ are as good as linear codes for q at least four
Approximations and Bounds for (n, k) Fork-Join Queues: A Linear Transformation Approach
Effective Edge-Fault-Tolerant Single-Source Spanners via Best (or Good) Swap Edges
Extremal copositive matrices with minimal zero supports of cardinality two
Deep Residual Learning for Weakly-Supervised Relation Extraction
Importance sampling for metastable and multiscale dynamical systems
Zero-temperature dynamics in the dilute Curie-Weiss model
Non-Count Symmetries in Boolean & Multi-Valued Prob. Graphical Models
Divisibility properties of the tangent numbers and its generalizations
Sequential Inverse Approximation of a Regularized Sample Covariance Matrix
The Garden of Eden theorem: old and new
Methods for compressible fluid simulation on GPUs using high-order finite differences
Providing Self-Aware Systems with Reflexivity
Estimating parameters of a directed weighted graph model with beta-distributed edge-weights
Transition to Chaos in the Kinetic Model of Cellulose Hydrolysis Under Enzyme Biosynthesis Control
Coloring ($P_5$, bull)-free graphs
Non-Coherent Detection for Diffusive Molecular Communications
Continuous-time statistics and generalized relaxation equations
Multi-critical behaviour of 4-dimensional tensor models up to order 6
Line codes generated by finite Coxeter groups
P-splines with an $\ell_1$ penalty for repeated measures
An Introduction to OFDM-OQAM
Anisotropic EM Segmentation by 3D Affinity Learning and Agglomeration
Concise Radiometric Calibration Using The Power of Ranking
Handwritten character recognition using some (anti)-diagonal structural features
Building Detection from Satellite Images on a Global Scale

Advertisements