arXiv research

The paper aims to develop new combinatorial dimensions for bounded memory learning.

Study reduces memory needs for active learning with enriched queries.

New framework captures long-term decision dependence in online learning.

Deep learning typically requires training a very capable architecture using large datasets. However, many important learning problems demand an ability to draw valid inferences from small size datasets, and such problems pose a particular challenge for deep learning. In this regard, various researches on "meta-learning…

New algorithms for multitask learning with long-term memory.

A new method stabilizes deep reinforcement learning by using QGraphs to retain replay memory information.

Memory-augmented neural networks consisting of a neural controller and an external memory have shown potentials in long-term sequential learning. Current RAM-like memory models maintain memory accessing every timesteps, thus they do not effectively leverage the short-term memory held in the controller. We hypothesize t…

We show that fundamental learning tasks, such as finding an approximate linear separator or linear regression, require memory at least \emph{quadratic} in the dimension, in a natural streaming setting. This implies that such problems cannot be solved (at least in this setting) by scalable memory-efficient streaming alg…

PySAD offers a unified Python framework for efficient streaming anomaly detection.

Algorithm learns linear systems from partial observations with near-optimal rate.

Study language generation with limited memory, showing different impacts on achievable densities and convergence.

We propose a new method to study the internal memory used by reinforcement learning policies. We estimate the amount of relevant past information by estimating mutual information between behavior histories and the current action of an agent. We perform this estimation in the passive setting, that is, we do not interven…

Quantum memory limits set by relativity theory.

New analysis tightens memory capacity of Hopfield models using spherical codes.

In this paper, we study the generalization properties of online learning based stochastic methods for supervised learning problems where the loss function is dependent on more than one training sample (e.g., metric learning, ranking). We present a generic decoupling technique that enables us to provide Rademacher compl…

Study shows how to balance memory and learning efficiency in continual learning.

Paper introduces a new framework to improve sample efficiency in POMDPs learning.

A central challenge faced by memory systems is the robust retrieval of a stored pattern in the presence of interference due to other stored patterns and noise. A theoretically well-founded solution to robust retrieval is given by attractor dynamics, which iteratively clean up patterns during recall. However, incorporat…

We derive upper and lower bounds for the policy regret of $T$-round online learning problems with graph-structured feedback, where the adversary is nonoblivious but assumed to have a bounded memory. We obtain upper bounds of $\widetilde O(T^{2/3})$ and $\widetilde O(T^{3/4})$ for strongly-observable and weakly-observab…

Recent studies on automatic neural architectures search have demonstrated significant performance, competitive to or even better than hand-crafted neural architectures. However, most of the existing network architecture tend to use residual, parallel structures and concatenation block between shallow and deep features …

Study shows memory needs grow with task sequence length in continual learning.

New algorithms for streaming bandits with limited memory.

Improved deep learning model deployment on tiny MCUs with mixed-precision quantization.

Efficient algorithm predicts unknown linear systems with long-term memory.

We study the power of different types of adaptive (nonoblivious) adversaries in the setting of prediction with expert advice, under both full-information and bandit feedback. We measure the player's performance using a new notion of regret, also known as policy regret, which better captures the adversary's adaptiveness…

The study analyzes prediction errors in systems with memory kernels, providing bounds and stability results.

Study non-oblivious adversarial bandits with delayed feedback and propose algorithms with improved regret bounds.

We propose a novel active learning framework for activity recognition using wearable sensors. Our work is unique in that it takes physical and cognitive limitations of the oracle into account when selecting sensor data to be annotated by the oracle. Our approach is inspired by human-beings' limited capacity to respond …

Study on learning strategies in adaptive Markov games with policy regret as metric.

LEM efficiently models long-term sequences with gradients.

We analyze computational limits of modern Hopfield models based on pattern norms.

Online learning algorithms are designed to learn even when their input is generated by an adversary. The widely-accepted formal definition of an online algorithm's ability to learn is the game-theoretic notion of regret. We argue that the standard definition of regret becomes inadequate if the adversary is allowed to a…

Memory-constrained algorithms need superlinear memory for efficient convex optimization.

Improved single-pass streaming MAB regret bound to O(K^1/3T^2/3).

Stable Hadamard Memory improves reinforcement learning by efficiently managing memory.

Generalizes memory and forecasting capacities for nonlinear recurrent networks with dependent inputs.

Survey on statistical inference under memory constraints.

Quadratic memory is essential for optimal convex optimization queries.

In this paper, we propose a dual memory structure for reinforcement learning algorithms with replay memory. The dual memory consists of a main memory that stores various data and a cache memory that manages the data and trains the reinforcement learning agent efficiently. Experimental results show that the dual memory …

Proposes a novel SAM operator for separate item and relational memories.

Gradient descent is efficient for solving feasibility problems with minimal memory and queries.

Transformers recall from long distributions with statistical guarantees.

We propose the first reduction-based approach to obtaining long-term memory guarantees for online learning in the sense of Bousquet and Warmuth, 2002, by reducing the problem to achieving typical switching regret. Specifically, for the classical expert problem with $K$ actions and $T$ rounds, using our framework we dev…

Stochastic Gradient Descent (SGD) is a fundamental algorithm in machine learning, representing the optimization backbone for training several classic models, from regression to neural networks. Given the recent practical focus on distributed machine learning, significant work has been dedicated to the convergence prope…

A new optimization method handles Euclidean bounds efficiently.

Global convergence of an online (stochastic) limited memory version of the Broyden-Fletcher- Goldfarb-Shanno (BFGS) quasi-Newton method for solving optimization problems with stochastic objectives that arise in large scale machine learning is established. Lower and upper bounds on the Hessian eigenvalues of the sample …

Paper studies statistical-computational trade-offs in tensor PCA and related problems.

A new RNN model tackles long-time dependencies with fast, invertible, and memory-efficient hidden states.