Whats new on arXiv

Generalized Dynamic Factor Models and Volatilities: Consistency, rates, and prediction intervals

Innovation Representation of Stochastic Processes with Application to Causal Inference

Typically, real-world stochastic processes are not easy to analyze. In this work we study the representation of any stochastic process as a memoryless innovation process triggering a dynamic system. We show that such a representation is always feasible for innovation processes taking values over a continuous set. However, the problem becomes more challenging when the alphabet size of the innovation is finite. In this case, we introduce both lossless and lossy frameworks, and provide closed-form solutions and practical algorithmic methods. In addition, we discuss the properties and uniqueness of our suggested approach. Finally, we show that the innovation representation problem has many applications. We focus our attention to Entropic Causal Inference, which has recently demonstrated promising performance, compared to alternative methods.

Evoplex: A platform for agent-based modeling on networks

Evoplex is a fast, robust and extensible platform for developing agent-based models and multi-agent systems on networks. Each agent is represented as a node and interacts with its neighbors, as defined by the network structure. Evoplex is ideal for modeling complex systems, for example in evolutionary game theory and computational social science. In Evoplex, the models are not coupled to the execution parameters or the visualization tools, and there is a user-friendly graphical user interface which makes it easy for all users, ranging from newcomers to experienced, to create, analyze, replicate and reproduce the experiments.

Please Stop Explaining Black Box Models for High Stakes Decisions

There are black box models now being used for high stakes decision-making throughout society. The practice of trying to explain black box models, rather than creating models that are interpretable in the first place, is likely to perpetuate bad practices and can potentially cause catastrophic harm to society. There is a way forward — it is to design models that are inherently interpretable.

Attentive Long Short-Term Preference Modeling for Personalized Product Search

E-commerce users may expect different products even for the same query, due to their diverse personal preferences. It is well-known that there are two types of preferences: long-term ones and short-term ones. The former refers to user’ inherent purchasing bias and evolves slowly. By contrast, the latter reflects users’ purchasing inclination in a relatively short period. They both affect users’ current purchasing intentions. However, few research efforts have been dedicated to jointly model them for the personalized product search. To this end, we propose a novel Attentive Long Short-Term Preference model, dubbed as ALSTP, for personalized product search. Our model adopts the neural networks approach to learn and integrate the long- and short-term user preferences with the current query for the personalized product search. In particular, two attention networks are designed to distinguish which factors in the short-term as well as long-term user preferences are more relevant to the current query. This unique design enables our model to capture users’ current search intentions more accurately. Our work is the first to apply attention mechanisms to integrate both long- and short-term user preferences with the given query for the personalized search. Extensive experiments over four Amazon product datasets show that our model significantly outperforms several state-of-the-art product search methods in terms of different evaluation metrics.

Rare Events Analysis in Stochastic Models for Bacterial Evolution

Radical shifts in the genetic composition of large cell populations are rare events with quite low probabilities, which direct numerical simulations generally fail to evaluate accurately. We develop a large deviations framework for a class of Markov chains modeling genetic evolution of bacteria such as E. coli. In particular, we develop the cost functional and a backward search algorithm for discrete-time Markov chains which describe daily evolution of histograms of bacterial populations.

Ontology Matching Techniques: A Gold Standard Model

Typically an ontology matching technique is a combination of much different type of matchers operating at various abstraction levels such as structure, semantic, syntax, instance etc. An ontology matching technique which employs matchers at all possible abstraction levels is expected to give, in general, best results in terms of precision, recall and F-measure due to improvement in matching opportunities and if we discount efficiency issues which may improve with better computing resources such as parallel processing. A gold standard ontology matching model is derived from a model classification of ontology matching techniques. A suitable metric is also defined based on gold standard ontology matching model. A review of various ontology matching techniques specified in recent research papers in the area was undertaken to categorize an ontology matching technique as per newly proposed gold standard model and a metric value for the whole group was computed. The results of the above study support proposed gold standard ontology matching model.

Bull Bear Balance: A Cluster Analysis of Socially Informed Financial Volatility

Using a method rooted in information theory, we present results that have identified a large set of stocks for which social media can be informative regarding financial volatility. By clustering stocks based on the joint feature sets of social and financial variables, our research provides an important contribution by characterizing the conditions in which social media signals can lead financial volatility. The results indicate that social media is most informative about financial market volatility when the ratio of bullish to bearish sentiment is high, even when the number of messages is low. The robustness of these findings is verified across 500 stocks from both NYSE and NASDAQ exchanges. The reported results are reproducible via an open-source library for social-financial analysis made freely available.

InstaNAS: Instance-aware Neural Architecture Search

Neural Architecture Search (NAS) aims at finding one ‘single’ architecture that achieves the best accuracy for a given task such as image recognition.In this paper, we study the instance-level variation,and demonstrate that instance-awareness is an important yet currently missing component of NAS. Based on this observation, we propose InstaNAS for searching toward instance-level architectures;the controller is trained to search and form a ‘distribution of architectures’ instead of a single final architecture. Then during the inference phase, the controller selects an architecture from the distribution, tailored for each unseen image to achieve both high accuracy and short latency. The experimental results show that InstaNAS reduces the inference latency without compromising classification accuracy. On average, InstaNAS achieves 48.9% latency reduction on CIFAR-10 and 40.2% latency reduction on CIFAR-100 with respect to MobileNetV2 architecture.

Control with Distributed Deep Reinforcement Learning: Learn a Better Policy

Distributed approach is a very effective method to improve training efficiency of reinforcement learning. In this paper, we propose a new heuristic distributed architecture for deep reinforcement learning (DRL) algorithm, in which a PSO based network update mechanism is adopted to speed up learning an optimal policy besides using multiple agents for parallel training. In this mechanism, the update of neural network of each agent is not only according to the training result of itself, but also affected by the optimal neural network of all agents. In order to verify the effectiveness of the proposed method, the proposed architecture is implemented on the Deep Q-Network algorithm (DQN) and the Deep Deterministic Policy Gradient algorithm (DDPG) to train several typical control problems. The training results show that the proposed method is effective.

Autonomous Systems — An Architectural Characterization

The concept of autonomy is key to the IoT vision promising increasing integration of smart services and systems minimizing human intervention. This vision challenges our capability to build complex open trustworthy autonomous systems. We lack a rigorous common semantic framework for autonomous systems. It is remarkable that the debate about autonomous vehicles focuses almost exclusively on AI and learning techniques while it ignores many other equally important autonomous system design issues. Autonomous systems involve agents and objects coordinated in some common environment so that their collective behavior meets a set of global goals. We propose a general computational model combining a system architecture model and an agent model. The architecture model allows expression of dynamic reconfigurable multi-mode coordination between components. The agent model consists of five interacting modules implementing each one a characteristic function: Perception, Reflection, Goal management, Planning and Self-adaptation. It determines a concept of autonomic complexity accounting for the specific difficulty to build autonomous systems. We emphasize that the main characteristic of autonomous systems is their ability to handle knowledge and adaptively respond to environment changes. We advocate that autonomy should be associated with functionality and not with specific techniques. Machine learning is essential for autonomy although it can meet only a small portion of the needs implied by autonomous system design. We conclude that autonomy is a kind of broad intelligence. Building trustworthy and optimal autonomous systems goes far beyond the AI challenge.

Bayesian Nonparametric Analysis of Multivariate Time Series: A Matrix Gamma Process Approach

While there is an increasing amount of literature about Bayesian time series analysis, only a few Bayesian nonparametric approaches to multivariate time series exist. Most methods rely on Whittle’s Likelihood, involving the second order structure of a stationary time series by means of its spectral density matrix. This is often modeled in terms of the Cholesky decomposition to ensure positive definiteness. However, asymptotic properties such as posterior consistency or posterior contraction rates are not known. A different idea is to model the spectral density matrix by means of random measures. This is in line with existing approaches for the univariate case, where the normalized spectral density is modeled similar to a probability density, e.g. with a Dirichlet process mixture of Beta densities. In this work, we present a related approach for multivariate time series, with matrix-valued mixture weights induced by a Hermitian positive definite Gamma process. The proposed procedure is shown to perform well for both simulated and real data. Posterior consistency and contraction rates are also established.

Non-deterministic inference using random set models: theory, approximation, and sampling method

A random set is a generalisation of a random variable, i.e. a set-valued random variable. The random set theory allows a unification of other uncertainty descriptions such as interval variable, mass belief function in Dempster-Shafer theory of evidence, possibility theory, and set of probability distributions. The aim of this work is to develop a non-deterministic inference framework, including theory, approximation and sampling method, that deals with the inverse problems in which uncertainty is represented using random sets. The proposed inference method yields the posterior random set based on the intersection of the prior and the measurement induced random sets. That inference method is an extension of Dempster’s rule of combination, and a generalisation of Bayesian inference as well. A direct evaluation of the posterior random set might be impractical. We approximate the posterior random set by a random discrete set whose domain is the set of samples generated using a proposed probability distribution. We use the capacity transform density function of the posterior random set for this proposed distribution. This function has a special property: it is the posterior density function yielded by Bayesian inference of the capacity transform density function of the prior random set. The samples of such proposed probability distribution can be directly obtained using the methods developed in the Bayesian inference framework. With this approximation method, the evaluation of the posterior random set becomes tractable.

Minimum reversion in multivariate time series

We propose a new multivariate time series model in which we assume that each component has a tendency to revert to the minimum of all components. Such a specification is useful to describe phenomena where each member in a population which is subjected to random noise mimics the behaviour of the best performing member. We show that the proposed dynamics generate co-integrated processes.We characterize the model’s asymptotic properties for the case of two populations and show a stabilizing effect on long term dynamics in simulation studies. An empirical study involving human survival data in different countries provides an example which confirms the occurrence of the phenomenon of reversion to the minimum in real data.

• Rough center manifolds• Joint Facade Registration and Segmentation for Urban Localization• An overview of deep learning in medical imaging focusing on MRI• Rigidity for zero sets of Gaussian entire functions• Distributed model independent algorithm for spacecraft synchronization under relative measurement bias• Doeblin Trees• Predicting Gender from Iris Texture May Be Harder Than It Seems• Ensemble of Multi-View Learning Classifiers for Cross-Domain Iris Presentation Attack Detection• The promises and pitfalls of Stochastic Gradient Langevin Dynamics• Parallel approach to sliding window sums• Asymptotic Task-Based Quantization with Application to Massive MIMO• SparseCast: Hybrid Digital-Analog Wireless Image Transmission Exploiting Frequency Domain Sparsity• Learning to discover and localize visual objects with open vocabulary• A Lower Bound of the Number of Threshold Functions in Terms of Combinatorial Flags on the Boolean Cube• Defensive alliance polynomial• Exponential Separation between Quantum Communication and Logarithm of Approximate Rank• Reinforced Cross-Modal Matching and Self-Supervised Imitation Learning for Vision-Language Navigation• Planning in Dynamic Environments with Conditional Autoregressive Models• WarpGAN: Automatic Caricature Generation• 50 Years of Test (Un)fairness: Lessons for Machine Learning• Inexact SARAH Algorithm for Stochastic Optimization• Sparse PCA from Sparse Linear Regression• The Anatomy of a Cryptocurrency Pump-and-Dump Scheme• On the cumulative Parisian ruin of multi-dimensional Brownian motion models• Real-Time Sleep Staging using Deep Learning on a Smartphone for a Wearable EEG• A Model-Based Reinforcement Learning Approach for a Rare Disease Diagnostic Task• Cooperation in the spatial prisoner’s dilemma game with probabilistic abstention• Recovery guarantees for polynomial approximation from dependent data with outliers• Variational End-to-End Navigation and Localization• A Survey of Mobile Computing for the Visually Impaired• Foreground Clustering for Joint Segmentation and Localization in Videos and Images• Reducing the irreducible uncertainty in return periods of 21st-century precipitation extremes• Artificial Retina Using A Hybrid Neural Network With Spatial Transform Capability• Save Our Spectrum: Contact-Free Human Sensing Using Single Carrier Radio• Canonical Duality Theory and Algorithm for Solving Bilevel Knapsack Problems with Applications• FilterReg: Robust and Efficient Probabilistic Point-Set Registration using Gaussian Filter and Twist Parameterization• One Shot Domain Adaptation for Person Re-Identification• Frequency Principle in Deep Learning with General Loss Functions and Its Potential Application• Regression calibration to correct correlated errors in outcome and exposure• Collaging on Internal Representations: An Intuitive Approach for Semantic Transfiguration• A Policy Gradient Method with Variance Reduction for Uplift Modeling• An Unknown Input Multi-Observer Approach for Estimation, Attack Isolation, and Control of LTI Systems under Actuator Attacks• FPTAS for barrier covering problem with equal circles in 2D• Temporal Convolutional Neural Network for the Classification of Satellite Image Time Series• LSICC: A Large Scale Informal Chinese Corpus• Improving Gated Recurrent Unit Based Acoustic Modeling with Batch Normalization and Enlarged Context• Ergodicity of the infinite swapping algorithm at low temperature• Multiscale geometric feature extraction for high-dimensional and non-Euclidean data with application• Testing Multivariate Scatter Parameter in Elliptical Model based on Forward Search Method• Bringing a Blurry Frame Alive at High Frame-Rate with an Event Camera• Phase-only Image Based Kernel Estimation for Single-image Blind Deblurring• Implanting Rational Knowledge into Distributed Representation at Morpheme Level• Tweedie Gradient Boosting for Extremely Unbalanced Zero-inflated Data• An identity for the odd double factorial• Finite Time Analysis of Vector Autoregressive Models under Linear Restrictions• Cross-domain Deep Feature Combination for Bird Species Classification with Audio-visual Data• IDD: A Dataset for Exploring Problems of Autonomous Navigation in Unconstrained Environments• A Hybrid Model for Role-related User Classification on Twitter• 3D-LaneNet: end-to-end 3D multiple lane detection• City-Scale Road Audit System using Deep Learning• Multi-task Learning over Graph Structures• Optimization of Battery Energy Storage to Improve Power System Oscillation Damping• A Survey on Joint Object Detection and Pose Estimation using Monocular Vision• Assessing the Value of Including Unimodality Information in Distributionally Robust Optimization Applied to Optimal Power Flow• Bayesian Weighted Mendelian Randomization for Causal Inference based on Summary Statistics• Wavelet-based and Fourier-based multivariate Whittle estimation: multiwave• A novel particle swarm optimizer with multi-stage transformation and genetic operation for VLSI routing• Attentioned Convolutional LSTM InpaintingNetwork for Anomaly Detection in Videos• Augmenting Robot Knowledge Consultants with Distributed Short Term Memory• Learning Latent Beliefs and Performing Epistemic Reasoning for Efficient and Meaningful Dialog Management• Brain-inspired robust delineation operator• Efficient volatility estimation in a two-factor model• MonoGRNet: A Geometric Reasoning Network for Monocular 3D Object Localization• The Coulomb gas, potential theory and phase transitions• Interacting reinforced stochastic processes: statistical inference based on the weighted empirical means• Generalised Differential Privacy for Text Document Processing• On the Structure of Higher Order Voronoi Cells• Region Based Extensive Response Index Pattern for Facial Expression Recognition• Balanced triangulations on few vertices and an implementation of cross-flips• Sound Approximation of Programs with Elementary Functions• Rejoinder for ‘Probabilistic Integration: A Role in Statistical Computation?’• Variational Autoencoders for New Physics Mining at the Large Hadron Collider• A Rule-based Kurdish Text Transliteration System• On finiteness and tails of perpetuities under a Lamperti-Kiu MAP• A New Standard for the Analysis and Design of Replication Studies• A bijection between self-conjugate and ordinary partitions and counting simultaneous cores as its application• Gapsets and numerical semigroups• Multi-hierarchical Independent Correlation Filters for Visual Tracking• A Differential Topological View of Challenges in Learning with Feedforward Neural Networks• Statistical Analysis of Multiple Antenna Strategies for Wireless Energy Transfer• Investigation of Nonlinear Communication Channel with Small Dispersion via Stochastic Correlator Approach• Finding a Bounded-Degree Expander Inside a Dense One• On the cost of essentially fair clusterings• On the Relationship Between Inference and Data Privacy in Decentralized IoT Networks• Universal Semi-Supervised Semantic Segmentation• A Multi-step Piecewise Linear Approximation Based Solution for Load Pick-up Problem in Electrical Distribution System• Hessian Riemannian gradient flows in convex programming• Quantum reservoir computing• Analysis of Roll-Calls in the European Parliament by Multiple Partitioning of Multiplex Signed Networks• Local order in binary Ge-Te glasses – an experimental study• Nonnegativity for hafnians of certain matrices• Matchable Image Retrieval by Learning from Surface Reconstruction• Estimating Causal Effects With Partial Covariates For Clinical Interpretability• Pair-wise Exchangeable Feature Extraction for Arbitrary Style Transfer• Tell me something my friends do not know: Diversity maximization in social networks• Unsupervised learning with sparse space-and-time autoencoders• Clustering Residential Electricity Load Curves via Community Detection in Network• Even and odd trees• Waveforms of molecular oscillations reveal circadian timekeeping mechanisms• Computing with Chemical Reaction Networks: A Tutorial• The Architecture of Mr. DLib’s Scientific Recommender-System API• ParsRec: A Novel Meta-Learning Approach to Recommending Bibliographic Reference Parsers• Decentralizing Multicell Beamforming via Deterministic Equivalents• Disentangling sources of influence in online social networks• Automatic segmentation of the Foveal Avascular Zone in ophthalmological OCT-A images• Robustness against the channel effect in pathological voice detection• A Convolutional Neural Network based Live Object Recognition System as Blind Aid• Existence and extinction in finite time for Stratonovich gradient noise porous media equations• Towards Machine Learning Prediction of Deep Brain Stimulation (DBS) Intra-operative Efficacy Maps• Spectra of Sparse Non-Hermitian Random Matrices• Combining neural and knowledge-based approaches to Named Entity Recognition in Polish• Planning with Learned Binarized Neural Network Transition Models in Factored State and Action Spaces• Sparse spectral estimation with missing and corrupted measurements• Deep Laplacian Pyramid Network for Text Images Super-Resolution• Optimal input potential functions in the interacting particle system method• Context-Aware Crowd Counting• Bayesian kernel machine causal mediation analysis• A Framework for Implementing Machine Learning on Omics Data• Enumerative properties of restricted words and compositions• Scan2Mesh: From Unstructured Range Scans to 3D Meshes• Analysis of large sparse graphs using regular decomposition of graph distance matrices• Sentence Encoding with Tree-constrained Relation Networks• Auctioning Electricity under Deep Renewable Integration using a Penalty for Shortfall• Identifying treatment effect heterogeneity in dose-finding trials using Bayesian hierarchical models• On the Performance of Backhaul Constrained Cell-Free Massive MIMO with Linear Receivers• Robust Cross-View Gait Identification with Evidence: A Discriminant Gait GAN (DiGGAN) Approach on 10000 People• ExpandNets: Exploiting Linear Redundancy to Train Small Networks• hynet: An Optimal Power Flow Framework for Hybrid AC/DC Power Systems• An optimized Parallel Failure-less Aho-Corasick algorithm for DNA sequence matching• Deep Ensemble Tensor Factorization for Longitudinal Patient Trajectories Classification• On restricted permutations of ${1,\ldots,n}$• Tracing in 2D to Reduce the Annotation Effort for 3D Deep Delineation• EVM analysis of an Interference Limited SIMO-SC System With Independent and Correlated Channels• Deep Network Interpolation for Continuous Imagery Effect Transition• Delocalization and continuous spectrum for ultrametric random operators• Unsupervised 3D Shape Learning from Image Collections in the Wild• Predicting Language Recovery after Stroke with Convolutional Networks on Stitched MRI• Scene Categorization from Contours: Medial Axis Based Salience Measures• Quantum Log-Approximate-Rank Conjecture is also False• The mean-field quantum Heisenberg ferromagnet via representation theory• Comparing with octopi• Higher-order Projected Power Iterations for Scalable Multi-Matching• Geometric Ergodicity of Affine Processes on Cones• Scalable graph-based individual named entity identification• Challenges in the Automatic Analysis of Students’ Diagnostic Reasoning• Similarity-preserving Image-image Domain Adaptation for Person Re-identification• A deep neural network predicts survival after heart imaging better than cardiologists• Leveraging Filter Correlations for Deep Model Compression• CLEAR: A Dataset for Compositional Language and Elementary Acoustic Reasoning• Phase Distribution Control of a Population of Oscillators• Low-Dose CT via Deep CNN with Skip Connection and Network in Network• Convolutional Neural Networks Deceived by Visual Illusions• Stacked Spatio-Temporal Graph Convolutional Networks for Action Segmentation• Grammar-based Representation and Identification of Dynamical Systems• The SNOW Theorem Revisited• HOGWILD!-Gibbs can be PanAccurate• Visual Entailment Task for Visually-Grounded Language Learning• Base-Stations Up in the Air: Multi-UAV Trajectory Control for Min-Rate Maximization in Uplink C-RAN• GAN Dissection: Visualizing and Understanding Generative Adversarial Networks• Divergence radii and the strong converse exponent of classical-quantum channel coding with constant compositions

Like this:

Like Loading…

Related