Develops deep jump learning for continuous treatment OPE.
problem Estimating mean outcomes under new treatment rules using historical data from different rules.
method Adaptive deep discretization of continuous treatment space using deep learning and multi-scale change point detection.
result Validated method through theoretical results, simulations, and real application to Warfarin Dosing.
Deep RBVFs improve continuous control in RL.
problem Challenges in finding optimal actions for continuous actions in RL.
method Introduced deep radial-basis value functions (RBVFs) for continuous control.
result RBF-DQN significantly outperforms value-function-only baselines and is competitive with actor-critic algorithms.
New insights into continual learning for deep models, showing convergence issues but local linear solutions.
problem Challenges in continual learning for homogeneous deep models.
method Sequential projections onto task margin sets, leveraging nonconvex projection theory.
result Local linear convergence under certain conditions for homogeneous deep networks.
This paper develops variational continual learning (VCL), a simple but general framework for continual learning that fuses online variational inference (VI) and recent advances in Monte Carlo VI for neural networks. The framework can successfully train both deep discriminative models and deep generative models in compl…
Continuous-time SGD converges under certain conditions, useful for deep learning.
problem Minimizing population expected loss in learning problems.
method Continuous-time approximation of stochastic gradient descent.
result Establishes sufficient conditions for convergence, applicable to overparametrized neural networks.
Proposes a continuous flow model to understand and control instability in gradient descent for deep learning.
problem Understanding and controlling the instability of gradient descent in deep learning.
method Introduces the Principal Flow (PF), a continuous time flow that approximates gradient descent dynamics.
result The PF captures divergent and oscillatory behaviors of gradient descent, including escaping local minima and saddle points.
Continual Learning is a learning paradigm where learning systems are trained with sequential or streaming tasks. Two notable directions among the recent advances in continual learning with neural networks are (i) variational Bayes based regularization by learning priors from previous tasks, and, (ii) learning the s…
Continual learning is hard for AI, leading to forgetting old knowledge.
problem Catastrophic forgetting in AI when learning new data.
method Review of continual learning in deep learning.
result Challenges and insights in continual learning.
This paper presents a method to improve continual learning stability and plasticity.
problem Balancing learning stability and plasticity in deep learning.
method Batch-level Experience Replay with Review approach.
result Achieved 1st place in all three scenarios of the CVPR 2020 CLVision challenge.
CLOPS improves deep learning for continuous physiological data.
problem Deep learning struggles with streaming, multi-sensor clinical data.
method CLOPS is a replay-based continual learning strategy.
result CLOPS outperforms state-of-the-art methods in continual learning scenarios.
Deep learning solves complex stochastic control with jumps.
problem Solving high-dimensional stochastic control tasks with jumps.
method Model-based approach using two neural networks, iteratively trained with objectives derived from the Hamilton-Jacobi-Bellman equation.
result Demonstrates effectiveness in solving complex high-dimensional stochastic control tasks.
New method creates coresets for deep neural networks efficiently.
problem Efficiently handling large data streams with limited resources.
method Cardinality-constrained bilevel optimization for deep neural networks.
result Demonstrated efficient generation of coresets for deep neural networks.
New method prevents deep learning forgetting past by remembering key examples.
problem Catastrophic forgetting in continual learning.
method Functional regularisation using Gaussian Process formulation.
result Achieves state-of-the-art performance on benchmarks.
Paper proposes SDRL to improve continual learning with less computational cost.
problem Catastrophic forgetting in continual learning.
method SDRL method that refines gradients from memorized samples to reduce gradient diversity.
result SDRL shows better performance than state-of-the-art methods on multiple benchmark tasks.
Deep Reinforcement Learning (DRL) has been applied to address a variety of cooperative multi-agent problems with either discrete action spaces or continuous action spaces. However, to the best of our knowledge, no previous work has ever succeeded in applying DRL to multi-agent problems with discrete-continuous hybrid (…
This work bridges continual learning, active learning, and open set recognition in deep neural networks.
problem Protecting previously acquired representations from catastrophic forgetting in deep neural networks.
method Surveying the literature and proposing a consolidated view to integrate open set recognition and active learning principles.
result Joint improvement in alleviating catastrophic forgetting, querying data, selecting task orders, and robust open world application.
Bayesian approach helps update deep models without forgetting past data.
problem Updating deep models sequentially without forgetting past data.
method Bayesian inference for continual learning.
result Bayesian approach enables updating model beliefs with new data.
Most existing deep reinforcement learning (DRL) frameworks consider either discrete action space or continuous action space solely. Motivated by applications in computer games, we consider the scenario with discrete-continuous hybrid action space. To handle hybrid action space, previous works either approximate the hyb…
Deep learning networks are approximated using dynamical systems theory.
problem Understanding the approximation capabilities of deep learning networks.
method Modeling deep residual networks as continuous-time dynamical systems and using approximation theories in Lp. result Established general sufficient conditions for universal approximation of deep residual networks.
EVCL combines VCL and EWC to prevent forgetting new tasks.
problem Preventing catastrophic forgetting in continual learning.
method Hybrid model integrating VCL and EWC.
result Consistently outperforms baselines in learning new tasks.
Researchers dissect Neural ODEs to understand their dynamics.
problem Understanding the inner workings of Neural ODEs.
method Developing continuous-depth formulation to clarify design choices.
result Clarified the influence of design choices on Neural ODE dynamics.
Efficiently learns deep factor graphs using Gaussian belief propagation.
problem Learning in deep factor graphs with efficient inference.
method Treats all relevant quantities as random variables, uses belief propagation for inference.
result Efficiently solves training and prediction problems in deep factor graphs with belief propagation.
The paper proves deep learning can be robust with certain loss functions.
problem The robustness of deep learning models under flawed data.
method Empirical-risk minimization with unbounded, Lipschitz-continuous loss functions.
result These loss functions provide efficient prediction under minimal data assumptions.
Bayesian deep learning improves deep learning's capabilities across diverse settings.
problem Overlooked metrics, tasks, and data types in deep learning.
method Revisits strengths of Bayesian deep learning and addresses challenges.
result Bayesian deep learning can elevate deep learning's capabilities across diverse settings.
Paper introduces Auto DeepVis to explain catastrophic forgetting in continual learning.
problem Catastrophic forgetting in continual learning of deep neural networks.
method Auto DeepVis and critical freezing techniques to address catastrophic forgetting.
result Critical freezing outperforms other methods on both past and future tasks.
Blog post discusses various implementations of Fisher Information for EWC in continual learning.
problem Improving Elastic Weight Consolidation (EWC) results by optimizing Fisher Information computation.
method Empirically compares different implementations of Fisher Information for EWC.
result Many reported EWC results can be improved by changing Fisher Information computation methods.
Deep Hedging learns optimal strategies for various risk levels.
problem Finding optimal hedging policies for diverse risk aversions.
method Continuous Reinforcement Learning with actor-critic algorithm.
result Demonstrated effectiveness in a stochastic volatility model.
Continuous-time optimal stopping solved with deep reinforcement learning
problem Optimal stopping problems in continuous time
method CARLOS (Continuous-time Adaptive Reinforcement Learning for Optimal Stopping)
result Higher prices than existing Bermudan solvers, approaching American upper bound
VSDN models sporadic time series with neural SDEs.
problem Modeling irregular and sparse time series data.
method Variational Bayesian method and neural SDEs.
result VSDNs outperform state-of-the-art models in prediction and interpolation.
Improved reinforcement learning with deep learning.
problem Extending MultiGrid Reinforcement Learning to work with deep learning.
method Combining potential-based reward shaping with a learned potential function from interaction, and adapting it for deep learning algorithms.
result DQN augmented with the approach performs significantly better on continuous control tasks.
This work compresses sequences by treating them as continuous-time processes, enabling efficient discretization.
problem Efficient compression of sequences, especially with deep learning models that scale with sequence length.
method Treat sequences as continuous-time processes, learn efficient discretization, and decode at different time intervals.
result Automatic bit rate reductions in video and motion capture sequences using learned discretization.
The feasibility of deep neural networks (DNNs) to address data stream problems still requires intensive study because of the static and offline nature of conventional deep learning approaches. A deep continual learning algorithm, namely autonomous deep learning (ADL), is proposed in this paper. Unlike traditional deep …
Develops DPG methods for continuous-time RL with deterministic policies.
problem High variance and slow convergence in stochastic policy RL methods.
method Derives continuous-time policy gradient formula and proposes CT-DDPG algorithm.
result CT-DDPG achieves superior stability and faster convergence in continuous-time RL.
Continual lifelong learning is essential to many applications. In this paper, we propose a simple but effective approach to continual deep learning. Our approach leverages the principles of deep model compression, critical weights selection, and progressive networks expansion. By enforcing their integration in an itera…
Generalization bounds derived for neural ODEs and deep residual networks.
problem Understanding the generalization capability of neural ODEs and deep residual networks.
method Lipschitz-based argument and analogy with deep residual networks.
result A generalization bound involving the magnitude of weight matrix differences.
AI learns to learn sequentially without forgetting.
problem Preventing catastrophic forgetting in machine learning models.
method Meta-learning a neuromodulatory activation-gating function to control selective activation in deep neural networks.
result State-of-the-art continual learning performance with 600 classes (9,000 updates).
Investigates gradient descent dynamics and introduces new regularisation methods.
problem Understanding and mitigating gradient descent instabilities and interactions with smoothness regularisation.
method Derives continuous-time flows to account for discretisation drift, constructs learning rate schedules and regularisers.
result New regularisation methods improve performance in reinforcement learning.
Dividing deep learning models for consistent anomaly detection in changing log data.
problem Anomaly detection methods fail when log data types change, leading to false negatives.
method Divide deep learning models based on log data correlation and extract correlations.
result Continues anomaly detection accuracy even when log data changes.
RANDPOL uses randomized networks for efficient reinforcement learning in continuous state and action MDPs.
problem Efficient reinforcement learning in environments with continuous state and action spaces.
method RANDPOL uses randomized function approximation to represent policy and value functions, providing finite time guarantees and improved numerical performance.
result RANDPOL achieves better numerical performance and provides finite time guarantees compared to deep neural network based algorithms.
Paper solves POMDPs in continuous time and discrete spaces.
problem Optimal decision making in discrete state and action space systems under partial observability.
method Combining optimal filtering theory and deep learning to solve a Hamilton-Jacobi-Bellman equation.
result Derives a mathematical description and solution approach for continuous-time POMDPs.
Efficient deep policy gradient method for continuous-time control problems.
problem Optimal control in continuous time with fine time discretization.
method Multi-scale deep policy gradient method with varying time discretization.
result Targeted efficiency in computational resources achieved through multi-scale approach.
Paper formalizes continual semi-supervised anomaly detection, showing promising results.
problem Formalizing continual semi-supervised anomaly detection in real-world conditions.
method Baseline model of variational autoencoder (VAE) with deep generative replay and outlier rejection.
result Outlier rejection shows promising results, often surpassing baseline methods.
A model combines GRU-D for missing data and Neural ODEs for time series continuity.
problem Challenges of informative missingness in multivariate time series data.
method Combines GRU-D for missing data imputation and Neural ODEs for temporal continuity.
result Demonstrates improved performance on a time series classification task.
A new trading model combines GARCH and PPO for better stock trading profits.
problem Limited performance of reinforcement learning in stock market trading.
method Parallel-network continuous trading model using GARCH and PPO.
result The model achieves more profit compared to traditional reinforcement learning methods.
Derives EoM for DNNs to describe GD dynamics precisely.
problem Gaps between differential equations and actual DNN learning dynamics due to discretization error.
method Starts from GF, derives counter term to cancel discretization error, obtains EoM.
result EoM precisely describes GD dynamics of DNNs, highlights differences between continuous and discrete GD.
UDENet and ResNet can approximate any function, with ODENet showing UAP for continuous functions.
problem Approximating any function using ODENet and ResNet.
method Proved UAP for ODENet and ResNet, derived gradient, and applied to various problems.
result UDENet and ResNet can approximate any function, with ODENet showing UAP for continuous functions.
Deep learning improves analysis of complex natural processes.
problem Simplistic dynamics in regression analyses of complex natural processes.
method Flexible function approximation using deep learning, relaxing standard assumptions.
result Substantial improvements in behavioral and neuroimaging data.
Current Deep Reinforcement Learning algorithms still heavily rely on handcrafted neural network architectures. We propose a novel approach to automatically find strong topologies for continuous control tasks while only adding a minor overhead in terms of interactions in the environment. To achieve this, we combine Neur…