Online Hyperparameter Optimization for Class-Incremental Learning
Class-incremental learning (CIL) aims to train a classification model while the number of classes increases phase-by-phase. An inherent challenge of CIL is the stability-plasticity tradeoff, i.e., CIL models should keep stable to retain old knowledge and keep plastic to absorb new knowledge. However, none of the existing CIL models can achieve the optimal tradeoff in different data-receiving settings--where typically the training-from-half (TFH) setting needs more stability, but the training-from-scratch (TFS) needs more plasticity. To this end, we design an online learning method that can adaptively optimize the tradeoff without knowing the setting as a priori. Specifically, we first introduce the key hyperparameters that influence the trade-off, e.g., knowledge distillation (KD) loss weights, learning rates, and classifier types. Then, we formulate the hyperparameter optimization process as an online Markov Decision Process (MDP) problem and propose a specific algorithm to solve it. We apply local estimated rewards and a classic bandit algorithm Exp3 to address the issues when applying online MDP methods to the CIL protocol. Our method consistently improves top-performing CIL methods in both TFH and TFS settings, e.g., boosting the average accuracy of TFH and TFS by 2.2 percentage points on ImageNet-Full, compared to the state-of-the-art.
Code (1)
Tasks
class-incremental learningClass Incremental LearningHyperparameter OptimizationIncremental LearningKnowledge DistillationMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Multilayer Neuromodulated Architectures for Memory-Constrained Online Continual Learning
We focus on the problem of how to achieve online continual learning under memory-constrained conditions where the input data may not be known \emph{a priori}. These constraints are relevant in edge computing scenarios. W…
Bayesian Optimizationclass-incremental learningClass Incremental LearningContinual Learning+4AdaCL:Adaptive Continual Learning
Class-Incremental Learning aims to update a deep classifier to learn new categories while maintaining or improving its accuracy on previously observed classes. Common methods to prevent forgetting previously learned clas…
Bayesian Optimizationclass-incremental learningClass Incremental LearningContinual Learning+1Dynamic Surrogate Switching: Sample-Efficient Search for Factorization Machine Configurations in Online Recommendations
Hyperparameter optimization is the process of identifying the appropriate hyperparameter configuration of a given machine learning model with regard to a given learning task. For smaller data sets, an exhaustive search i…
Hyperparameter OptimizationSubset Sampling For Progressive Neural Network Learning
Progressive Neural Network Learning is a class of algorithms that incrementally construct the network's topology and optimize its parameters based on the training data. While this approach exempts the users from the manu…
Face RecognitionIncremental Search Space Construction for Machine Learning Pipeline Synthesis
Automated machine learning (AutoML) aims for constructing machine learning (ML) pipelines automatically. Many studies have investigated efficient methods for algorithm selection and hyperparameter optimization. However, …
AutoMLBIG-bench Machine LearningHyperparameter Optimization