Volume-preserving neural networks prevent gradient issues.
problem Vanishing and exploding gradients in deep neural networks.
method A new neural network architecture with volume-preserving sublayers.
result Volume-preserving neural networks maintain gradient stability.
Active learning holds promise of significantly reducing data annotation costs while maintaining reasonable model performance. However, it requires sending data to annotators for labeling. This presents a possible privacy leak when the training set includes sensitive user data. In this paper, we describe an approach for…
Tempered sigmoids improve deep learning privacy.
problem Privacy-preserving deep learning with strict differential privacy guarantees.
method Developed tempered sigmoid activation functions for deep learning models.
result Tempered sigmoids outperform ReLU in achieving state-of-the-art accuracy.
Proposes privacy-preserving sensor data transformations to prevent user re-identification and sensitive activity inference.
problem Privacy threats from shared sensor data and potential user re-identification.
method Mechanisms to transform sensor data to eliminate patterns for re-identification and sensitive activity inference, while maintaining minor utility loss.
result Reduced user re-identification accuracy to random guess level and prevented inference of sensitive activities.
IMPACT optimizes LLM compression by focusing on activation importance, reducing model size up to 55.4%.
problem Resource constraints in deploying large language models (LLMs).
method IMPACT integrates activation importance into low-rank compression, optimizing for both size and accuracy.
result IMPACT achieves up to 55.4% greater model size reduction while maintaining comparable or better accuracy.
This paper converts NACE classification into embeddings to preserve hierarchical structure.
problem Preserving hierarchical structure in NACE classification while reducing dimensions.
method Custom metrics for hierarchical structure retention; state-of-the-art models and dimensionality reduction.
result The proposed approach effectively preserves hierarchical structures in NACE classification.
Binary autoencoder with sparse hidden layer preserves information and zero reconstruction error.
problem Preserving information and zero reconstruction error in binary neural networks.
method Binary autoencoder with random binary weights, sparse hidden layer, and varying neuron thresholds.
result Zero reconstruction error for any input with a large hidden layer and varying neuron thresholds.
State-of-the-art subspace clustering methods are based on expressing each data point as a linear combination of other data points while regularizing the matrix of coefficients with ℓ1, ℓ2 or nuclear norms. ℓ1 regularization is guaranteed to give a subspace-preserving affinity (i.e., there are no conne…
New initialization schemes preserve fractional moments of weights in deep networks, improving training and test performance.
problem Heavy-tailed distribution of stochastic gradients in DNNs during training.
method Developed initialization schemes that preserve any given fractional moment of order s < 2 over layers for various activations.
result The network output admits a heavy-tailed distribution with finite moments, improving training and test performance.
Traditionally, the automatic recognition of human activities is performed with supervised learning algorithms on limited sets of specific activities. This work proposes to recognize recurrent activity patterns, called routines, instead of precisely defined activities. The modeling of routines is defined as a metric lea…
Improved diffusion models for image synthesis with better training dynamics.
problem Uneven and ineffective training in diffusion models.
method Redesigned network layers to preserve activation, weight, and update magnitudes.
result Significantly better networks at equal computational complexity, improving FID to 1.81.
Active learning improves GP regression on complex, high-dimensional data.
problem Improving Gaussian Process regression in high-dimensional spaces with discontinuous functions.
method Combines manifold learning with active learning to optimize data selection and reduce dimensionality.
result Superior performance over random learning in synthetic data experiments.
Deep Learning techniques have achieved remarkable results in many domains. Often, training deep learning models requires large datasets, which may require sensitive information to be uploaded to the cloud to accelerate training. To adequately protect sensitive information, we propose distributed layer-partitioned train…
Training neural networks under a strict Lipschitz constraint is useful for provable adversarial robustness, generalization bounds, interpretable gradients, and Wasserstein distance estimation. By the composition property of Lipschitz functions, it suffices to ensure that each individual affine transformation or nonline…
GAPA method provides efficient uncertainty quantification for pretrained networks.
problem Reliable uncertainty estimates for pretrained models are challenging.
method Post-hoc Gaussian Process Activations (GAPA) method that shifts Bayesian modeling from weights to activations.
result GAPA method provides efficient uncertainty quantification without altering the backbone's predictions.
Pruning neural networks adds differential privacy noise, preserving data utility.
problem Achieving differential privacy in neural networks without sacrificing data utility.
method Proving equivalence between pruning and adding differential privacy noise to hidden-layer activations.
result Pruning can be a more effective alternative to adding differential privacy noise for neural networks.
We present a novel factor analysis method that can be applied to the discovery of common factors shared among trajectories in multivariate time series data. These factors satisfy a precedence-ordering property: certain factors are recruited only after some other factors are activated. Precedence-ordering arise in appli…
Active inference framework improves U-statistic estimation efficiency.
problem Costly acquisition of labels for U-statistics. method Active inference framework with optimal sampling rule.
result Substantial gains in estimation efficiency over baseline methods.
New method preserves privacy while improving machine learning accuracy.
problem Privacy-preserving machine learning for daily data.
method Compressive Privacy and multi-kernel method.
result Improved utility classification accuracy with privacy preservation.
Framework tackles class imbalance and noisy labels in active learning.
problem Class imbalance and noisy labels in real-world datasets.
method Uses foundation model priors to select informative samples for active learning.
result Substantial annotation savings (over 50%) with preserved performance and robustness.
BiTAT improves neural network quantization for edge devices by focusing on weight dependencies and disentangling them.
problem Performance degradation of compact neural networks under extreme quantization.
method Task-dependent Aggregated Transformation (BiTAT) method that orthonormalizes weights and progressively quantizes them.
result BiTAT effectively preserves model performance on ImageNet and CIFAR-100 with compact backbones.
Unified framework recovers exact input from SOM activation patterns.
problem Generating high-dimensional data from Self-Organizing Maps (SOMs).
method Inverting SOM activation patterns to recover input, using linear system and Tikhonov regularization.
result MUSIC framework produces coherent semantic transitions and maintains high classifier confidence.
In the last few years there has been a growing interest in Human Activity Recognition~(HAR) topic. Sensor-based HAR approaches, in particular, has been gaining more popularity owing to their privacy preserving nature. Furthermore, due to the widespread accessibility of the internet, a broad range of streaming-based app…
Study develops a new model for predicting individual mobility based on activity patterns.
problem Existing mobility prediction methods overlook underlying activity patterns.
method Activity-based hidden Markov model (IOHMM) using transit smart card data.
result Proposed model achieves similar prediction performance to LSTM but also provides behavioral insights.
An increasing number of sensors on mobile, Internet of things (IoT), and wearable devices generate time-series measurements of physical activities. Though access to the sensory data is critical to the success of many beneficial applications such as health monitoring or activity recognition, a wide range of potentially …
ADS filters data points for efficient batch active learning.
problem Efficiently selecting data points for annotation in parallel settings.
method Active Data Shapley (ADS) using the Shapley value of data.
result Significantly increases efficiency of active learning by 6x.
Standard neural network architectures are non-linear only by virtue of a simple element-wise activation function, making them both brittle and excessively large. In this paper, we consider methods for making the feed-forward layer more flexible while preserving its basic structure. We develop simple drop-in replacement…
FAQ efficiently evaluates LLMs with statistical guarantees using adaptive query selection.
problem Efficiently evaluating many LLMs on a large suite of benchmarks is expensive.
method FAQ uses Bayesian factor models, adaptive sampling, and proactive active inference to select queries.
result FAQ delivers up to 5x effective sample size gains over baselines, matching CI width with fewer queries.
Study efficient interactive learning for structured outputs with reliable computation.
problem Interactive learning with noisy labels and structured output spaces.
method Identify and utilize CRISPs (probabilistic models) that guarantee reliable and efficient computation of probabilistic quantities.
result CRISPs enable robust and efficient active and skeptical learning in large structured output spaces.
Union of Subspaces (UoS) is a popular model to describe the underlying low-dimensional structure of data. The fine details of UoS structure can be described in terms of canonical angles (also known as principal angles) between subspaces, which is a well-known characterization for relative subspace positions. In this pa…
Data-aware activation function customization reduces neural network error.
problem Current neural networks lack consideration for specific activation functions.
method Linear algebraic explanation and Diaconis-Shahshahani Approximation Theorem criteria for activation functions.
result Using an even activation function like seagull can reduce neural network error by orders of magnitude.
This paper proposes grid cells encode position via a conformal isometric embedding of 2D physical space.
problem Hexagonal grid firing patterns in grid cells.
method Learning a distance-preserving position embedding in neural space using a recurrent neural network.
result The conformal isometric embedding of 2D physical space into neural space explains hexagonal grid firing patterns.
PPAT uses predictions to improve risk estimation in active testing.
problem Exploiting informative predictions from black-box models for efficient risk estimation.
method Combines LURE estimator with prediction-powered control variate.
result PPAT outperforms existing methods in risk estimation and uncertainty quantification.
Neural networks are known to be vulnerable to adversarial examples. Carefully chosen perturbations to real images, while imperceptible to humans, induce misclassification and threaten the reliability of deep learning systems in the wild. To guard against adversarial examples, we take inspiration from game theory and ca…
Before training a neural net, a classic rule of thumb is to randomly initialize the weights so the variance of activations is preserved across layers. This is traditionally interpreted using the total variance due to randomness in both weights \emph{and} samples. Alternatively, one can interpret the rule of thumb as pr…
We introduce a conceptually simple and scalable framework for continual learning domains where tasks are learned sequentially. Our method is constant in the number of parameters and is designed to preserve performance on previously encountered tasks while accelerating learning progress on subsequent problems. This is a…
SympNets identify Hamiltonian systems from data using linear, activation, and gradient modules.
problem Identifying Hamiltonian systems from data.
method Composition of linear, activation, and gradient modules; universal approximation theorems.
result SympNets can approximate arbitrary symplectic maps and generalize well to various Hamiltonian systems.
A major challenge in computational chemistry is the generation of novel molecular structures with desirable pharmacological and physiochemical properties. In this work, we investigate the potential use of autoencoder, a deep learning methodology, for de novo molecular design. Various generative autoencoders were used t…
New method improves ResNet performance without batch normalization.
problem Improving ResNet performance without batch normalization.
method Adapted Weight Standardization to maintain signal propagation.
result Highly performant ResNets achieve state-of-the-art performance on ImageNet.
Post-hoc uncertainty quantification improves on pre-trained neural networks without underfitting.
problem Uncertainty quantification in neural networks is underfitting or computationally demanding.
method Gaussian Process Activation function (GAPA) for neuron-level uncertainty, with two methods: GAPA-Free and GAPA-Variational.
result GAPA-Variational outperforms Laplace approximation on most datasets in uncertainty quantification metrics.
Improved fraud detection in finance with quantum-enhanced federated learning.
problem Challenges in detecting financial fraud with traditional methods.
method Hybrid quantum-enhanced federated learning framework combining quantum LSTM with privacy-preserving techniques.
result Approximately 5% improvement in performance metrics compared to conventional models.
It has been noted in existing literature that over-parameterization in ReLU networks generally improves performance. While there could be several factors involved behind this, we prove some desirable theoretical properties at initialization which may be enjoyed by ReLU networks. Specifically, it is known that He initia…
Classification algorithms aim to predict an unknown label (e.g., a quality class) for a new instance (e.g., a product). Therefore, training samples (instances and labels) are used to deduct classification hypotheses. Often, it is relatively easy to capture instances but the acquisition of the corresponding labels remai…
Paper proposes active learning for sound event detection with reduced annotation effort.
problem Reducing annotation effort for sound event detection.
method Change point detection for candidate selection, mismatch-first farthest-traversal for selection, training with context recordings.
result The proposed system achieves similar performance to full annotation with only 2% of data, reducing annotation effort.
Paper connects MoE and self-attention, proposing active-attention.
problem Improving efficiency and performance of self-attention mechanisms.
method Established connection between MoE and self-attention, analyzed quadratic gating functions, proposed active-attention mechanism.
result Active-attention outperforms standard self-attention in various tasks.
We consider a few quantities that characterize trading on a stock market in a fixed time interval: logarithmic returns, volatility, trading activity (i.e., the number of transactions), and volume traded. We search for the power-law cross-correlations among these quantities aggregated over different time units from 1 mi…
A new deep learning framework selects representative samples for unsupervised learning.
problem Selecting representative samples for unsupervised learning in non-linear data.
method DUAL framework using an encoder-decoder architecture to learn nonlinear embeddings and a selection block to choose representative samples.
result DUAL outperforms state-of-the-art methods in selecting representative samples for unsupervised learning.
Theory explains deep nonlinear networks' plateaus and transitions.
problem Understanding long plateaus and feature acquisition transitions in deep nonlinear networks.
method Derived an exact identity for Frobenius norms, classified activation functions, and reduced matrix flow to a scalar ODE.
result Escape time law τ⋆=Θ(ε−(r−2)) for deep nonlinear networks, where r is the number of bottleneck layers.