Aims to describe neural network training dynamics using two-time-scale models.
problem Lack of a general mathematical description of neural network training.
method Introduces a theoretical framework based on two-time-scale population dynamics.
result Derives selection-mutation equations and effective fitness for hyperparameters.
Fiber simplifies RL and population-based methods for distributed training.
problem Challenges in RL and population-based methods, including frequent interaction with simulations and dynamic scaling.
method Introducing Fiber, a scalable distributed computing framework.
result Significantly expands accessibility of large-scale parallel computation.
FIRE PBT improves neural network training by focusing on long-term performance.
problem Greedy decision mechanisms in PBT lead to poor long-term performance.
method FIRE PBT uses a fitness metric to encourage long-term performance over short-term improvements.
result FIRE PBT outperforms PBT on ImageNet and matches hand-tuned learning rates.
Framework improves policy generalizability under biased training data.
problem Learning policies that generalize to a target population from biased training data.
method Characterizes sample selection bias using a selection variable, optimizes minimax value over uncertainty set, derives efficient algorithm.
result Policies generalize to target population, outperform standard methods.
Copula-based method generates synthetic populations from marginal distributions.
problem Generating realistic synthetic populations from limited data.
method Copula-based framework for population synthesis.
result Copula framework enhances transferability and realism of synthetic populations.
PB2 uses probabilistic models to efficiently discover high-performing hyperparameters.
problem Efficiently tuning hyperparameters in reinforcement learning (RL) is computationally expensive and laborious.
method Population-Based Bandits (PB2) uses probabilistic models to guide hyperparameter exploration.
result PB2 discovers high-performing hyperparameter configurations with fewer agents than PBT.
Robust RL improves controller robustness to dynamics variations using adversarial populations.
problem Robustness issues in RL when dynamics are perturbed.
method Adversarial population augmentation to the Robust RL formulation.
result Population-based adversarial approach yields more robust and generalizable policies.
New method learns population dynamics from snapshots using JKO scheme and inverse optimization.
problem Recovering underlying process governing particle evolution from discrete time samples.
method Combines JKO scheme with inverse optimization techniques for end-to-end adversarial training.
result Improved performance over prior JKO-based methods with theoretical guarantees.
A new model synthesizes population with fewer structural and sampling zeros.
problem Synthesizing a feasible and diverse synthetic population from limited data.
method A deep generative model with two regularizations to minimize structural zeros and preserve sampling zeros.
result The model significantly improves feasibility and diversity of synthetic populations.
This work learns models for population dynamics using variational methods and higher-order quadrature.
problem Modeling population dynamics of physical systems with stochastic and mean-field effects.
method Variational problem to infer gradient fields, combining Monte Carlo sampling with higher-order quadrature rules.
result Accurate prediction of population dynamics over a wide range of parameters.
Despite recent innovations in network architectures and loss functions, training RNNs to learn long-term dependencies remains difficult due to challenges with gradient-based optimisation methods. Inspired by the success of Deep Neuroevolution in reinforcement learning (Such et al. 2017), we explore the use of gradient-…
Develops robust learning methods for datasets with sub-populations.
problem Robust performance and generalization to unseen testing populations in datasets with sub-populations.
method Min-max-regret (MMR) formulation for distribution-free robust hierarchical model.
result Empirical MMR enjoys regret guarantees on training and unseen testing populations.
A framework is presented for unsupervised learning of representations based on infomax principle for large-scale neural populations. We use an asymptotic approximation to the Shannon's mutual information for a large neural population to demonstrate that a good initial approximation to the global information-theoretic o…
New bounds for KANs trained with DP-SGD, addressing correlated noise.
problem Risk bounds for Kolmogorov-Arnold Networks trained by DP-SGD with correlated noise.
method Established new optimization and population risk analysis for KANs trained with DP-SGD, addressing correlated noise.
result First optimization and population risk analysis of correlated-noise mechanisms for DP training in non-convex settings, including neural networks.
New method pools labels from similar data items to improve learning from small samples.
problem Learning from small, human-annotated samples with potential disagreement among annotators.
method Proposes neighborhood-based pooling for sharing labels across similar data items.
result Improves learning from small, noisy samples by pooling labels from similar items.
Improves exploration in reinforcement learning with diverse population.
problem Lack of diversity in reinforcement learning agents leads to limited exploration.
method Optimizes a population of agents using volume of behavioral manifold and online learning techniques.
result Improves exploration without sacrificing performance.
NeuPL learns diverse policies in strategy games efficiently.
problem Iterative training of policies in strategy games leads to under-trained good-responses and wasteful repetition.
method NeuPL uses a single conditional model to represent a population of policies, offering convergence guarantees and transfer learning.
result NeuPL achieves better performance and efficiency across various domains, enabling access to novel strategies.
Dynamic meta-learning improves multi-agent communication with natural language.
problem Static population-based approaches struggle with adapting to human partners using natural language.
method Dynamic population-based meta-learning approach.
result Agents outperform all prior work in communicating with seen partners and humans.
NPE trains neural networks to approximate posterior distributions in SIR models from final outcome data.
problem Computational challenges in Bayesian inference for SIR models with final outcome data.
method Neural posterior estimation (NPE) using a logNormal posterior approximated by a neural network.
result NPE accurately recovers reference posteriors across various population sizes and transmission regimes.
DICE learns population dynamics from discrete samples.
problem Learning smooth population dynamics from discrete data.
method Discrete Inverse Continuity Equation (DICE) method.
result DICE models are stable and generate representative samples.
GNMC reduces XCSF population size while preserving function approximation and policy accuracy.
problem Population bloat in XCSF compaction.
method Introduced GNMC, a novel compaction algorithm.
result GNMC reduces population size significantly without compromising function approximation or policy accuracy.
Automates RL with sample-efficient hyperparameter optimization.
problem Challenges in applying deep RL due to hyperparameter sensitivity and inefficiency.
method Population-based AutoRL framework for meta-optimizing RL algorithms and architectures.
result Reduces the number of environment interactions needed for meta-optimization by up to an order of magnitude.
MFM integrates multiple evolving populations using Wasserstein manifold flows.
problem Learning dynamics of multiple interacting populations evolving over time.
method Meta Flow Matching (MFM) integrates vector fields on Wasserstein manifold using amortized flow models and GNN embeddings.
result MFM improves prediction of individual treatment responses on multi-patient single-cell drug screen data.
SaML guides ML models to avoid survey biases.
problem ML models trained on survey data often ignore survey design metadata.
method Nine-step guideline for incorporating survey design metadata in ML lifecycle.
result SaML provides valid population inference from survey data.
New method bypasses global fit for LISA's Galactic binaries, extracting population parameters directly.
problem Disentangling LISA's Galactic binary sources from backgrounds in a computationally intensive process.
method Simulation-based approach using normalizing flow to infer population parameters.
result Direct inference of population parameters from LISA's frequency strain series.
EPC curriculum improves MARL performance as agent population grows.
problem Challenges in learning good policies for large multi-agent systems.
method Evolutionary Population Curriculum (EPC) for scaling MARL.
result EPC consistently outperforms baselines as agent population increases.
ScoreFusion fuses multiple diffusion models to enhance generative modeling of a target population.
problem Enhancing generative modeling of a target population with limited data.
method ScoreFusion uses KL barycenters of auxiliary populations and recasts the learning problem as score matching in denoising diffusion.
result ScoreFusion achieves a dimension-free sample complexity bound in total variation distance.
Analyzes how bias evolves in SGD training across different data sub-populations.
problem Understanding bias formation during machine learning training.
method Analytical description of SGD dynamics in a teacher-student setup with Gaussian-mixture model.
result Different sub-populations influence bias at different timescales, revealing shifting classifier preferences.
Infinitesimal boosting converges to a deterministic process in large sample limit.
problem Characterizing the asymptotic behavior of infinitesimal gradient boosting in large sample sizes.
method Proving convergence to a deterministic process using large sample theory and differential equations.
result The test error decreases over time in the population limit.
High-throughput 3D control training system achieves 100,000 FPS.
problem Lack of efficient, single-machine reinforcement learning systems.
method Sample Factory combines asynchronous sampling and off-policy correction.
result Achieves 100,000 FPS on 3D control problems without sacrificing sample efficiency.
RECaST calibrates source models for target populations with uncertainty quantification.
problem Uncertainty in transfer learning predictions.
method Random effect calibration of source to target models.
result Nominal coverage of prediction sets in linear models, robust to nonlinear approximations.
Estimates Mozambique's population using remote sensing and microcensus data.
problem Lack of frequent population estimation due to censuses lacking spatio-temporal resolution.
method Combines remote sensing, microcensus data, and transfer learning with publicly available datasets.
result Population predictions improve with footprint area estimation using transfer learning.
Improved disentanglement of data factors using recursive training.
problem Current unsupervised disentanglement methods are inconsistent and fail to achieve levels of disentanglement seen in supervised approaches.
method Introduced PBT for VAEs, used UDR for heuristic scoring, and developed recursive rPU-VAE approach.
result Recursive training leads to robust disentanglement of data factors across multiple datasets.
Estimates personalized policies robust to shifts in target populations.
problem Estimating policies that perform well in diverse target populations.
method Develops methods for estimating robust policies considering shifts in outcomes and characteristics.
result Welfare-maximizing policies are robust to certain shifts in potential outcomes.
Databases of electronic health records (EHRs) are increasingly used to inform clinical decisions. Machine learning methods can find patterns in EHRs that are predictive of future adverse outcomes. However, statistical models may be built upon patterns of health-seeking behavior that vary across patient subpopulations, …
ABPS improves RL training efficiency by sharing policies and evolving hyper-params.
problem Data inefficiency in training deep RL models for real-world applications.
method ABPS: adaptive behavior policy sharing; ABPS-PBT: hybridizing ABPS with PBT for evolving hyper-params.
result ABPS achieves superior performance and reduced variance compared to conventional hyper-parameter tuning.
Study proves convergence of subgradients for optimal transport-based objectives.
problem Ensuring statistical consistency and optimization stability in transport-based models.
method Proves graphical convergence of subdifferentials to the subdifferential of the population objective.
result Standard subgradient methods consistently approach stationary points of the population-level problem.
Proposes CRA framework for certifying fair predictive models.
problem Certifying fairness of predictive models trained on biased data.
method Formulates CRA for fairness queries, uses background knowledge and limited target population stats.
result Builds certifiably fair predictive models on target population.
New framework for online control in evolving populations.
problem Control of evolving populations in real-world conditions.
method Online control framework for linear and non-linear dynamical systems.
result Near-optimal regret bounds for gradient-based controllers.
The paper learns pose variations within shape populations using constrained mixtures of factor analyzers.
problem Learning pose variations within a shape population with articulated parts and relative rotations.
method Formulated as mixtures of factor analyzers, segmentation by component posterior probabilities, and constraints on factor loading matrices for rotation matrices.
result Automatic learning of pose variations from shape populations, resulting in smooth and realistic animations.
Many cooperative multiagent reinforcement learning environments provide agents with a sparse team-based reward, as well as a dense agent-specific reward that incentivizes learning basic skills. Training policies solely on the team-based reward is often difficult due to its sparsity. Furthermore, relying solely on the a…
The paper analyzes the maximum margin algorithm's performance on noisy data.
problem Analyzing the performance of maximum margin algorithm on noisy data.
method Finite-sample analysis of maximum margin algorithm applied to noisy data.
result The maximum margin algorithm can achieve nearly optimal population risk with sufficient over-parameterization.
In this paper, we will provide an introduction to the derivative-free optimization algorithms which can be potentially applied to train deep learning models. Existing deep learning model training is mostly based on the back propagation algorithm, which updates the model variables layers by layers with the gradient desc…
The quantification problem consists of determining the prevalence of a given label in a target population. However, one often has access to the labels in a sample from the training population but not in the target population. A common assumption in this situation is that of prior probability shift, that is, once the la…
In order to investigate the breast cancer prediction problem on the aging population with the grades of DCIS, we conduct a tree augmented naive Bayesian network experiment trained and tested on a large clinical dataset including consecutive diagnostic mammography examinations, consequent biopsy outcomes and related can…
Paper develops a two-population model to assess longevity basis risk.
problem Mismatch between hedger's liability and hedging instrument causes longevity basis risk.
method Develops a two-population mortality model using Lee-Carter model and renewal process.
result Proposed model provides significant risk reduction when mortality jumps and sampling risk are considered.
This paper studies the landscape of empirical risk of deep neural networks by theoretically analyzing its convergence behavior to the population risk as well as its stationary points and properties. For an l-layer linear neural network, we prove its empirical risk uniformly converges to its population risk at the rat…
Exploiting the wealth of imaging and non-imaging information for disease prediction tasks requires models capable of representing, at the same time, individual features as well as data associations between subjects from potentially large populations. Graphs provide a natural framework for such tasks, yet previous graph…