Mixture of basis for interpretable continual learning with distribution shifts

by   Mengda Xu, et al.

Continual learning in environments with shifting data distributions is a challenging problem with several real-world applications. In this paper we consider settings in which the data distribution(task) shifts abruptly and the timing of these shifts are not known. Furthermore, we consider a semi-supervised task-agnostic setting in which the learning algorithm has access to both task-segmented and unsegmented data for offline training. We propose a novel approach called mixture of Basismodels (MoB) for addressing this problem setting. The core idea is to learn a small set of basis models and to construct a dynamic, task-dependent mixture of the models to predict for the current task. We also propose a new methodology to detect observations that are out-of-distribution with respect to the existing basis models and to instantiate new models as needed. We test our approach in multiple domains and show that it attains better prediction error than existing methods in most cases while using fewer models than other multiple model approaches. Moreover, we analyze the latent task representations learned by MoB and show that similar tasks tend to cluster in the latent space and that the latent representation shifts at the task boundaries when tasks are dissimilar.


page 7

page 8

page 14

page 15

page 16

page 17

page 18


Task-agnostic Continual Hippocampus Segmentation for Smooth Population Shifts

Most continual learning methods are validated in settings where task bou...

Online Continual Learning with Natural Distribution Shifts: An Empirical Study with Visual Data

Continual learning is the problem of learning and retaining knowledge th...

SIESTA: Efficient Online Continual Learning with Sleep

In supervised continual learning, a deep neural network (DNN) is updated...

Assessing Robustness of EEG Representations under Data-shifts via Latent Space and Uncertainty Analysis

The recent availability of large datasets in bio-medicine has inspired t...

Real World Time Series Benchmark Datasets with Distribution Shifts: Global Crude Oil Price and Volatility

The scarcity of task-labeled time-series benchmarks in the financial dom...

Variational Beam Search for Online Learning with Distribution Shifts

We consider the problem of online learning in the presence of sudden dis...

Learned Indexes for Dynamic Workloads

The recent proposal of learned index structures opens up a new perspecti...

Please sign up or login with your details

Forgot password? Click here to reset