Abstract:Deep neural network ensembles combine the wisdom of multiple deep neural networks to improve the generalizability and robustness over individual networks. It has gained increasing popularity to study deep ensemble techniques in the deep learning community. Some mission-critical applications utilize a large number of deep neural networks to form deep ensembles to achieve desired accuracy and resilience, which introduces high time and space costs for ensemble execution. However, it still remains a critical challenge whether a small subset of the entire deep ensemble can achieve the same or better generalizability and how to effectively identify these small deep ensembles for improving the space and time efficiency of ensemble execution. This paper presents a novel deep ensemble pruning approach, which can efficiently identify smaller deep ensembles and provide higher ensemble accuracy than the entire deep ensemble of a large number of member networks. Our hierarchical ensemble pruning approach (HQ) leverages three novel ensemble pruning techniques. First, we show that the focal diversity metrics can accurately capture the complementary capacity of the member networks of an ensemble, which can guide ensemble pruning. Second, we design a focal diversity based hierarchical pruning approach, which will iteratively find high quality deep ensembles with low cost and high accuracy. Third, we develop a focal diversity consensus method to integrate multiple focal diversity metrics to refine ensemble pruning results, where smaller deep ensembles can be effectively identified to offer high accuracy, high robustness and high efficiency. Evaluated using popular benchmark datasets, we demonstrate that the proposed hierarchical ensemble pruning approach can effectively identify high quality deep ensembles with better generalizability while being more time and space efficient in ensemble decision making.

A pruning algorithm for training neural network ensembles

Class-Aware Pruning for Efficient Neural Networks

A Pruning Algorithm for Training Cooperative Neural Network Ensembles

A Probabilistic Ensemble Pruning Algorithm

Hierarchical Pruning of Deep Ensembles with Focal Diversity

Statistical Mechanical Analysis of Neural Network Pruning

Ensemble Pruning Via Constrained Eigen-Optimization

Prune and Tune Ensembles: Low-Cost Ensemble Learning with Sparse Independent Subnetworks

An optimal pruning algorithm of classifier ensembles: dynamic programming approach

A Novel Bayesian Ensemble Pruning Method

Pareto Ensemble Pruning

Margin distribution and structural diversity guided ensemble pruning

Data Dependant Learners Ensemble Pruning

Predictive Ensemble Pruning by Expectation Propagation

Margin-Based Pareto Ensemble Pruning: an Ensemble Pruning Algorithm That Learns to Search Optimized Ensembles.

Emerging Paradigms of Neural Network Pruning

Subarchitecture Ensemble Pruning in Neural Architecture Search

(Pen-) Ultimate DNN Pruning

Sparse optimization guided pruning for neural networks

Learning Sparse Structured Ensembles with Stochastic Gradient MCMC Sampling and Network Pruning.

A Lempel-Ziv Complexity-Based Neural Network Pruning Algorithm