Paper tackles stutter detection using deep learning.
problem Identification and classification of stuttered speech.
method Uses a deep residual network with bidirectional LSTM layers.
result Achieves an average miss rate of 10.03%, outperforming state-of-the-art.
The paper argues for applying fairness in machine learning, even partial, as an improvement.
problem The lack of applied fairness in machine learning products.
method Elaborates on the importance of applying fairness, even partial, in machine learning systems.
result The paper supports the idea that even partial fairness is better than no fairness.
New argument suggests torsion cannot be part of gravity models.
problem The presence of torsion in gravity models is debated.
method Used spectral geometry and pseudo-differential calculus.
result No well-defined functional for torsion in spectral formulation.
Discrete theory for rational maps improved by generalized branch points.
problem Discretization effects in locating branch points in circle packings.
method Introducing generalized branch points that can be positioned anywhere in the geometry.
result Fixed flaws in discrete Ahlfors and Weierstrasse functions using generalized branching.
Synthetic datasets help study bias in ML, overcoming data scarcity.
problem Lack of relevant datasets for bias research in ML.
method Presented a family of synthetic datasets with adjustable bias levels.
result Demonstrated an experiment using synthetic data to study bias.
Develops workflow for synthetic insurance datasets.
problem Lack of realistic publicly available insurance datasets.
method Uses CTGAN neural network architecture to generate tabular data.
result Synthesized datasets evaluated positively in multiple aspects.
Paper tackles overestimation bias in continuous control, improving performance by 25%.
problem Overestimation bias in off-policy learning.
method Truncated Quantile Critics (TQC) combines distributional representation, truncation, and ensembling of critics.
result TQC outperforms state-of-the-art methods by 25% on the Humanoid environment.
EdgeAI aims to deploy deep learning on IoT devices.
problem Deep learning's high computational demands on IoT devices.
method Addressing data-independent deployment and communication-aware distributed inference.
result New directions to enable deep learning on IoT devices.
Machine learning lacks causal models, hindering strong AI.
problem Current machine learning systems lack a model of reality.
method Demonstrates seven tasks beyond current ML capabilities using causal inference.
result Machine learning needs causal models for strong AI.
Explains how machine learning models can be biased and presents interactive plots to visualize bias.
problem Bias in machine learning models can lead to unfair decisions.
method Develops interactive plots to visualize bias in machine learning models.
result Demonstrates that machine learning models can learn and propagate bias from the data.
The correspondence of stationary, axisymmetric, asymptotically flat space-times and bundles over a reduced twistor space has been established in four dimensions. The main impediment for an application of this correspondence to examples in higher dimensions is the lack of a higher-dimensional equivalent of the Ernst pot…
FlexServe simplifies deployment of PyTorch models as REST endpoints.
problem Lack of control over model evolution and strict security requirements in operational environments.
method Developed FlexServe, a library to deploy multi-model ensembles with flexible batching.
result Rapid deployment of PyTorch models without intermediate transformations.
Develops a method to define admissible rewards for robust policy evaluation in RL.
problem Defining a reward function for robust off-policy evaluation in RL with limited data.
method Identifies an admissible set of reward functions ensuring policies are close to past behavior and can be evaluated with high confidence.
result Demonstrates the approach on synthetic and real-world domains, including a critical care application.
Randomized algorithms that base iteration-level decisions on samples from some pool are ubiquitous in machine learning and optimization. Examples include stochastic gradient descent and randomized coordinate descent. This paper makes progress at theoretically evaluating the difference in performance between sampling wi…
New method aligns brain surfaces based on functional signatures.
problem Inter-individual variability in neuroimaging data.
method Fused Unbalanced Gromov-Wasserstein (FUGW) based on Optimal Transport.
result FUGW significantly increases between-subject correlation of activity.
New method circumvents non-convexity in bilevel RL via hyper-gradient.
problem Non-convexity in lower-level RL problems in bilevel reinforcement learning.
method Characterizing hyper-gradient via fully first-order information, circumventing convexity assumption.
result Developed model-based and model-free algorithms with convergence rate O(ε−1). Infinite-horizon Gaussian processes reduce computational complexity for long datasets.
problem Cubic computational cost in state dimensionality for Gaussian processes.
method Single-sweep EP inference scheme for GPs with general likelihoods, reducing cost to O(m^2) per data point.
result Reduced computational complexity from cubic to quadratic in state dimensionality.
The paper revisits classical competition theory to explain speculative asset price dynamics.
problem Understanding the dynamics of speculative asset prices and their volatility.
method Specialized classical model of competition with reservation prices, incorporating speculation.
result The model explains excess, fat-tailed, and clustered volatility in speculative asset prices.
New framework aims to make neural network explanations more reliable.
problem Current interpretability methods rely on intuition and lack falsifiability.
method Proposes a framework for strongly falsifiable interpretability research.
result Falsifiable interpretability methods can generate meaningful advances in understanding DNNs.
Bayesian approach improves crowdsourcing predictions.
problem Estimating continuous labels from crowdsource workers.
method Variational Bayesian technique for worker noise models.
result Bayesian approaches significantly outperform non-Bayesian methods.
Paper develops a new similarity metric for predicting stock market returns.
problem Predicting stock returns is challenging due to market stochasticity and various influencing factors.
method Case-based reasoning approach using historical pricing data and a novel similarity metric.
result Demonstrates the benefits of the novel similarity metric in predicting stock market returns.
NeuroFabric proposes a method to optimize sparse network training topologies.
problem Long training times in deep neural networks due to high memory and compute requirements.
method Developed a new sparse neural network initialization scheme and evaluated various topologies.
result Identified a single optimal topology that maximizes accuracy across different datasets.
Polynomial time algorithm for learning mixtures of Mallows models with any number of components.
problem Learning parameters of mixtures of Mallows models with any constant number of components.
method Determinantal identity of Zagier, polynomial identifiability, test functions, information-theoretic lower bounds, local queries, beyond worst-case analysis.
result First polynomial time algorithm for provably learning mixtures of Mallows models with any constant number of components.
Improved ICU mortality prediction with interpretable deep learning.
problem Sub-optimal performance of traditional mortality prediction scores.
method Deep multi-scale convolutional architecture trained on MIMIC-III, coalitional game theory for visual explanations.
result State-of-the-art performance with interpretability.
The study examines machine learning classification algorithms and their generalizability using Framingham Heart Study data.
problem Addressing biases and generalizability issues in machine learning classification algorithms.
method Comparison of eight machine learning classification algorithms on Framingham Heart Study data.
result Double discriminant scoring of type I is the most generalizable algorithm.
ReNA fast clusters features for structured signals, reducing analysis time and noise.
problem Efficiently summarize structured signals to reduce analysis time and noise.
method Recursive Nearest Agglomeration (ReNA) for linear-time feature clustering.
result ReNA approximates data as well as traditional methods but with linear time complexity.
CARP speeds up convex clustering by 100x and offers better visualization.
problem Computational intensity and lack of compelling visualizations in convex clustering.
method Algorithmic Regularization for iterative approximation of regularization paths.
result CARP delivers over 100-fold speed-up and finer approximation grid.
This paper is concerned with offline reinforcement learning (RL), which learns using pre-collected data without further exploration. Effective offline RL would be able to accommodate distribution shift and limited data coverage. However, prior algorithms or analyses either suffer from suboptimal sample complexities or …
Picket guards against corrupted data in machine learning models.
problem Data corruption biases models and invalidates predictions.
method PicketNet detects corrupted data using self-supervised deep learning; flags corrupted queries online.
result Picket consistently protects models from corrupted data during training and deployment.
Enhances Deep Hedging with K-FAC for financial data.
problem High computational burden in training neural networks for financial applications.
method Integrates Kronecker-Factored Approximate Curvature (K-FAC) optimization with LSTM networks.
result Significant improvements in convergence and hedging efficacy, reducing transaction costs and P&L variance.
FastDeepIoT optimizes neural network execution time on mobile devices.
problem Excessive neural network execution time on low-end devices.
method Automatically learns execution time model and optimizes network configurations.
result Significant reduction in execution time and energy consumption.
New method removes interference bias in causal models.
problem Interference bias impedes causal effect identification in real-world settings.
method Novel definition of causal models with local interference, semi-parametric assumptions.
result True Average Causal Effect can be identified in certain semi-parametric models with local interference.
Paper proposes a new method for uncertainty estimation in medical data.
problem Difficulty in assigning confidence to deep learning model predictions in healthcare.
method Combines deep Bayesian learning with deep kernel learning for uncertainty estimation.
result Demonstrates improved uncertainty estimation compared to Gaussian processes and deep Bayesian neural networks.
Bayesian model estimates ACT impact on pediatric AML survival.
problem Estimating ACT impact on survival in AML patients with treatment timing issues.
method Generative Bayesian semi-parametric model with Gamma Process priors.
result Posterior inference for ACT efficacy under dynamic rules.
Efficiently infers graph edges from genetic similarity data in landscape genetics.
problem Inferring unknown graph edges from genetic similarity data in a heterogeneous landscape.
method Developed an efficient first-order optimization method to solve the inverse landscape genetics problem.
result Our method provides fast and reliable convergence, significantly outperforming existing heuristics.
High-dimensional neural network manifolds misalign with human perception, causing adversarial examples.
problem Adversarial attacks fool neural networks, but their origin is unclear.
method Defined and analyzed a network's perceptual manifold (PM) for a class concept.
result Neural network PMs have orders of magnitude higher dimensions than natural human concepts, suggesting exponential misalignment.
New method reduces mobile device energy for image classification.
problem High energy consumption of CNNs on mobile devices.
method Global optimization of CNN architectures with Bayesian optimization.
result Reduces energy consumption by up to 6x compared to blackbox CNNs.
Continued reliance on human operators for managing data centers is a major impediment for them from ever reaching extreme dimensions. Large computer systems in general, and data centers in particular, will ultimately be managed using predictive computational and executable models obtained through data-science tools, an…
Deep learning uses WiFi CSI for reliable human presence detection.
problem Reliable human presence detection using ambient RF signals.
method Pre-processing of estimated CSI followed by deep learning.
result Near perfect presence detection during multiple extended periods.
Bayesian UQ matches frequentist UQ for adaptively collected data.
problem Uncertainty quantification for adaptive data collection.
method Extends Bernstein-von Mises theorem to adaptively collected data.
result Bayesian UQ asymptotically matches Wald-type frequentist UQ.
Improves Gaussian Process for scalable modeling of large datasets.
problem Scalability issues in Gaussian Process modeling for large datasets.
method Adaptive Sequential Monte Carlo (ASMC) for training Gaussian Process (GP).
result ASMC implementation enables accurate and scalable modeling of large-scale industry problems.
Develops a method to efficiently learn causal DAGs using directed clique trees.
problem Efficiently learning causal DAGs in the presence of large cliques.
method Decomposes DAGs into independently orientable components using directed clique trees and designs a two-phase intervention algorithm.
result Proves that the number of single-node interventions necessary to orient any DAG in an EC is at least the sum of half the size of the largest cliques in each chain component of the essential graph.
TrialGraph uses graph machine learning to improve clinical trial design and predict side effects.
problem Complexity and cost in clinical trials hinder drug development.
method Curated clinical trial data set converted to graph-structured formats, applied graph machine learning algorithms.
result MetaPath2Vec algorithm performed exceptionally well, improving prediction accuracy.
A new machine learning model forecasts COVID-19 incidence at county level in the USA.
problem Inaccurate disease spread forecasting due to spatiotemporal homogeneity assumptions.
method Spatiotemporal machine learning using LSTM architecture with spatial and temporal features.
result COVID-LSTM outperforms COVID-19 Forecast Hub's Ensemble model in accuracy.
This paper optimizes how many samples are needed to estimate a population's binary responses.
problem Estimating a distribution from incomplete or corrupted samples.
method The approach involves computing the empirical mean of a certain function, pre-solving a linear program, and using complex-analytic methods.
result Optimal sample complexity for population recovery is determined, showing phase transitions and sensitivity to dimension.
The abstract covers various aspects of eBusiness and eGovernment, including digital currencies, m-government services, gender inclusivity, eLearning, export performance, SME digitalization, and banking customer behavior.
problem Various challenges and opportunities in eBusiness and eGovernment.
method Critical review, UTAUT model with perceived risk theory, GAD approach, inductive research paradigm, one-on-one interviews, survey questionnaires, convenience sampling.
result Impediments to eLearning uptake, gender inclusivity in e-procurement, export performance of manufacturing firms, SME digitalization impact, measuring and modeling framework for Internet banking customers.