Improving Fine-Tuning with Latent Cluster Correction
Abstract
The formation of salient clusters in the latent spaces of a neural network (NN) during training strongly impacts its final accuracy on classification tasks. This paper proposes a novel fine-tuning method that boosts performance by improving the quality of these latent clusters, using the Louvain community detection algorithm and a specifically designed loss function. We present preliminary results that demonstrate that this process yields an appreciable accuracy gain for classical NN architectures fine-tuned on the CIFAR100 dataset.
Cite
Text
Thanh. "Improving Fine-Tuning with Latent Cluster Correction." NeurIPS 2024 Workshops: FITML, 2024.Markdown
[Thanh. "Improving Fine-Tuning with Latent Cluster Correction." NeurIPS 2024 Workshops: FITML, 2024.](https://mlanthology.org/neuripsw/2024/thanh2024neuripsw-improving/)BibTeX
@inproceedings{thanh2024neuripsw-improving,
title = {{Improving Fine-Tuning with Latent Cluster Correction}},
author = {Thanh, Cédric Ho},
booktitle = {NeurIPS 2024 Workshops: FITML},
year = {2024},
url = {https://mlanthology.org/neuripsw/2024/thanh2024neuripsw-improving/}
}