Probabilistic descent on manifolds using triangular sets and embedding theorems.
problem Descent over manifolds defined by polynomials.
method Triangularization, embedding theorem, numerical continuation.
result Effective numerical method for probabilistic descent.
Stein's method improves probabilistic inference and learning.
problem Improving probabilistic inference and learning methods.
method Constructing Stein discrepancies from Stein operators and Stein sets, discussing their properties.
result Connection between Stein operators and Stein variational gradient descent.
Proposes VSGD optimizer combining probabilistic and gradient-based methods.
problem Uncertainty modeling in deep neural networks.
method Combines probabilistic and gradient-based approaches using SVI.
result VSGD outperforms Adam and SGD on image classification tasks.
MT-SGD samples from multiple target distributions using gradient descent.
problem Sampling from multiple unnormalized target distributions.
method Proposes MT-SGD, a flow of intermediate distributions to sample from multiple target distributions.
result Asymptotic analysis shows MT-SGD reduces to multiple-gradient descent for multi-objective optimization.
Probabilistic models with deep neural networks now handle large data sets.
problem Historical constraints on probabilistic modeling.
method Variational inference, stochastic gradient descent, distributed computation.
result Probabilistic models can now handle large data sets and complex relationships.
Variational inference approximates the posterior distribution of a probabilistic model with a parameterized density by maximizing a lower bound for the model evidence. Modern solutions fit a flexible approximation with stochastic gradient descent, using Monte Carlo approximation for the gradients. This enables variatio…
In deterministic optimization, line searches are a standard tool ensuring stability and efficiency. Where only stochastic gradients are available, no direct equivalent has so far been formulated, because uncertain gradients do not allow for a strict sequence of decisions collapsing the search space. We construct a prob…
In deterministic optimization, line searches are a standard tool ensuring stability and efficiency. Where only stochastic gradients are available, no direct equivalent has so far been formulated, because uncertain gradients do not allow for a strict sequence of decisions collapsing the search space. We construct a prob…
New method classifies manifold-valued data using Riemannian geometry.
problem Classifying data on curved Riemannian manifolds.
method Probabilistic Learning Vector Quantization on Symmetric Positive Definite Matrices.
result The method outperforms traditional Euclidean methods on manifold-valued data.
This work links SOMs and GMMs, providing a mathematical basis for their use.
problem Understanding the relationship between SOMs and GMMs.
method Mathematical treatment showing SOMs as gradient descent on a GMM log-likelihood.
result SOMs can be interpreted as probabilistic models, justifying their use in various applications.
Study Markov chain gradient descent in Hilbert spaces for quadratic loss.
problem Approximating optimal solutions for quadratic loss functions.
method Developed a Markov chain-based stochastic gradient algorithm in Hilbert spaces.
result Established probabilistic upper bounds on convergence.
Study on gradient descent in Hilbert spaces with Markov chains, focusing on mixing coefficients.
problem Analyzing convergence of gradient descent in Hilbert spaces with stationary Markov chains.
method Examined strictly stationary Markov chains with φ- and β-mixing coefficients, derived probabilistic upper bounds. result Probabilistic upper bounds on convergence behavior of gradient descent algorithm based on mixing coefficients.
Probabilistic models predict neural network performance across varying hyperparameters.
problem Predicting neural network performance with different hyperparameters.
method Probabilistic models based on random forests and Bayesian recurrent neural networks.
result Models outperform state-of-the-art hyperparameter optimization methods.
Scalable solver reduces PDE uncertainty with active learning.
problem High computational cost in solving PDEs.
method Stochastic dual descent and clustering-based active learning.
result Solver scales to large number of collocation points.
Stochastic gradient methods enable learning probabilistic models from large amounts of data. While large step-sizes (learning rates) have shown to be best for least-squares (e.g., Gaussian noise) once combined with parameter averaging, these are not leading to convergent algorithms in general. In this paper, we conside…
New method uses backward SDEs for deep learning uncertainty.
problem Uncertainty quantification in deep learning models.
method Probabilistic machine learning with stochastic neural networks and stochastic optimal control.
result Effectiveness validated through numerical experiments.
Analyzes deep neural networks training errors with SGD and random init.
problem Lack of rigorous understanding of deep learning algorithms.
method Mathematical analysis of deep learning with SGD and random init.
result First full error analysis for deep learning with SGD and random init.
We extend quantum probabilistic models and develop a learning algorithm.
problem Characterize expressiveness and learn hidden quantum Markov models.
method Show HQMMs are a subclass of OOMs without negative probabilities. Develop a feasible gradient descent algorithm on the Stiefel manifold.
result Our learning algorithm is faster and scales better than previous methods.
Paper introduces probabilistic approach to CO layers in ML.
problem Challenges in training CO layers with stochastic gradient descent.
method Probabilistic perspective, approximate differentiation, structured losses.
result InferOpt.jl package for training CO layers.
Proposes a method to control model complexity in neural network optimization.
problem Reduces the computational cost of neural architecture search.
method Probabilistic model-based dynamic optimization with a penalty term to control model complexity.
result The proposed method controls model complexity while maintaining performance.
Stochastic gradient descent based algorithms are typically used as the general optimization tools for most deep learning models. A Restricted Boltzmann Machine (RBM) is a probabilistic generative model that can be stacked to construct deep architectures. For RBM with Bernoulli inputs, non-Euclidean algorithm such as st…
New method tracks time-varying parameters in data.
problem Tracking unknown time-varying parameters in data.
method Stochastic gradient descent-based recursive scheme with log-likelihood as gain function.
result Convergence in mean-square error in a suitable neighborhood of the unknown parameter.
Continuous optimization is an important problem in many areas of AI, including vision, robotics, probabilistic inference, and machine learning. Unfortunately, most real-world optimization problems are nonconvex, causing standard convex techniques to find only local optima, even with extensions like random restarts and …
Gradient descent on DDPM objective learns Gaussian mixtures efficiently.
problem Learning Gaussian mixtures using gradient-based methods.
method Gradient descent on DDPM objective, connecting to EM and spectral methods.
result Gradient descent can efficiently recover Gaussian mixture parameters under certain conditions.
We present sparse topical coding (STC), a non-probabilistic formulation of topic models for discovering latent representations of large collections of data. Unlike probabilistic topic models, STC relaxes the normalization constraint of admixture proportions and the constraint of defining a normalized likelihood functio…
Quantum algorithms solve structured prediction problems faster.
problem Solving structured prediction problems efficiently.
method Two quantum algorithms: one using quantum minimum finding, the other using Gibbs sampling.
result Quantum algorithms achieve a runtime scaling with the square root of the label space size and precision.
EM algorithm converges in KL divergence for exponential families via mirror descent.
problem Lack of understanding of EM's non-asymptotic convergence properties.
method Viewing EM as a mirror descent algorithm, showing convergence rates in KL divergence.
result KL divergence rates for EM in exponential families, invariant to parametrization.
Adaptive probabilistic load forecasting improves performance in power systems.
problem Complexity of electricity load forecasting due to changing drivers and local generation.
method Adaptive probabilistic approach using Kalman filter and online gradient descent.
result Adaptive probabilistic forecasts improve performance in both point and probabilistic forecasting.
Neural Processes combine GP flexibility with NN efficiency.
problem Combining the strengths of Gaussian Processes and Neural Networks.
method Combines Gaussian Processes' probabilistic modeling with Neural Networks' efficiency.
result Demonstrates improved performance on regression and optimization tasks.
New algorithm estimates complex probabilistic models efficiently.
problem Estimating complex probabilistic models with large data sets.
method Backfitting algorithm based on stochastic gradient descent.
result Outperforms other implementations for structured additive distributional regression.
Researchers study heavy-tail properties of SGD using stochastic recurrence equations.
problem Analyzing heavy-tail properties of Stochastic Gradient Descent (SGD).
method Modeling SGD iterations as multivariate affine stochastic recursions and applying the theory of irreducible-proximal (i-p) matrices.
result Extended results of Gürbüzbalaban et al. (2020) by using the theory of i-p matrices.
ErasureHead speeds up distributed GD with approximate gradient coding.
problem Mitigating delays in distributed gradient descent.
method Approximate gradient coding to tolerate delays.
result ErasureHead converges as quickly as GD and has faster runtime under probabilistic delays.
Quantum kernel improves probabilistic time series forecasting.
problem Quantifying uncertainty in probabilistic time series predictions.
method Integrates quantum kernel with Gaussian process regression.
result Quantum kernel enhances forecasting performance.
Gradient descent converges linearly for neural networks with specific conditions.
problem Optimizing neural networks with fixed width and depth.
method Local Polyak-Lojasiewicz criterion for gradient flow and descent.
result Gradient descent converges to zero-loss solutions under certain conditions.
Meta-learning for few-shot learning entails acquiring a prior over previous tasks and experiences, such that new tasks be learned from small amounts of data. However, a critical challenge in few-shot learning is task ambiguity: even when a powerful prior can be meta-learned from a large number of prior tasks, a small d…
This work improves understanding of dimension reduction algorithms and their probabilistic embeddings.
problem Improving theoretical understanding of non-linear dimension reduction algorithms.
method Analytical investigation of a generalized multidimensional scaling optimization problem.
result Probabilistic formulation of the problem leads to deterministic embeddings, contrary to standard implementations.
EKI solves machine learning tasks without gradients.
problem Machine learning tasks, especially those involving deep neural networks.
method Ensemble Kalman inversion (EKI) as a gradient-free algorithm.
result Wide applicability and robustness of EKI demonstrated.
New bounds enable training of probabilistic models for deep networks.
problem Training scalable latent variable models for deep networks.
method Introducing new variational bounds for specific output layers of neural networks.
result Analytical bounds for certain output layers allow training without re-parameterization or Monte Carlo approximations.
Efficiently samples and learns densities with symmetries using equivariant methods.
problem Efficiently sampling and learning densities with symmetries.
method Equivariant Stein Variational Gradient Descent (SVGD) and equivariant energy based models.
result Improves and scales up training of energy based models.
New algorithm improves efficiency of quantum system modeling.
problem Intractable complexities in quantum Hamiltonian learning and Gibbs sampling.
method Generalized quantum natural gradient descent and Quantum-Probabilistic Mirror Descent.
result Data sample efficiency proven using information geometry and quantum metrology.
Randomized SINDy learns dynamic data structures using probabilistic methods.
problem Learning time-dependent data structures in dynamic systems.
method Sequential machine learning with a probabilistic approach, incorporating feature augmentation and Tikhonov regularization.
result Demonstrated effectiveness in regression and binary classification using real-world data.
Algorithm improves convergence in stochastic optimization problems.
problem Efficiently constructing pre-conditioners for noisy stochastic optimization.
method Iterative probabilistic inference algorithm using a Gaussian noise model.
result Empirically demonstrates improved convergence in stochastic optimization problems.
New implicit Krasulina's k-PCA update avoids QR-decomposition and improves convergence.
problem Online k-PCA problem with orthonormality constraint.
method Derived an implicit form of Krasulina's update that bypasses orthonormality constraint.
result The new update avoids costly QR-decomposition and yields superior convergence.
The paper proposes a method to train time-varying generative models using natural gradients.
problem Training time-varying generative models efficiently and accurately.
method Projecting generative model parameters onto an exponential family manifold and optimizing using natural gradient descent.
result The proposed method efficiently approximates the natural gradient and can be applied to various exponential family models.
Stein Variational Gradient Descent optimizes particle sets to match distribution expectations.
problem Efficiently approximating complex distributions in machine learning.
method Evolve particle sets to match the expectations of a given distribution using Stein operators and kernels.
result Particles can be used to exactly estimate expectations of functions on distributions, providing insights into kernel choice.
Bayesian regularizations are explicitly implemented in CNNs, improving deep learning generalization.
problem Improving generalization in deep learning models.
method Introduced a novel probabilistic representation for CNN hidden layers and demonstrated their Bayesian nature.
result CNNs have explicitly Bayesian regularizations based on Bayesian regularization theory.
A new method learns graph compression from data.
problem Graphs lack ordering, making conventional compression algorithms ineffective.
method Partition and Code framework: decompose, learn, encode.
result PnC achieves compression gains that grow with graph size.
Bayesian optimisation for dynamically adjusting learning rates in machine learning models.
problem Dynamic adjustment of learning rates schedules in machine learning models.
method Probabilistic model based on latent Gaussian processes and auto-/regressive formulation.
result Flexibly adjusts learning rates schedules to abrupt changes of behaviours.