Evolutionary Instance Selection With Multiple Partial Adaptive Classifiers for Domain Adaptation

Bach Hoai Nguyen, Bing Xue, Peter Andreae, Mengjie Zhang

Abstract: Domain adaptation reuses the knowledge learned from an existing (source) domain to classify unlabeled data from another related (target) domain. However, the two domains have different data distributions. Common approaches to bridge the two distributions are selecting/reweighting instances, building domain-invariant feature subspaces, or directly building adaptive classifiers. Recent domain adaptation work has shown that combining the above first two approaches before applying the third approach achieves better performance than performing each approach individually. However, most existing instance selection approaches are based on a ranking mechanism, ignore interdependences between instances, and require a predefined number of selected instances. Furthermore, adaptive classifiers are sensitive to their parameters which are challenging to optimize due to the lack of target labeled instances. This article introduces a novel evolutionary instance selection approach for domain adaptation. We propose a compacted representation and an efficient fitness function for particle swarm optimization to automatically determine the number of selected instances while considering the interdependencies among instances. This article also proposes to use multiple partial classifiers to build a more reliable and robust adaptive classifier. The results show that evolutionary instance selection selects better instances than the ranking approach. In cooperation with multiple partial classifiers, the proposed algorithm achieves better performance than nine state-of-the-art and well-known domain adaptation approaches.

External IDs:doi:10.1109/tevc.2023.3346406