arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2207.04874cs.NEcs.LG
收藏

赫布连续表示学习

Hebbian Continual Representation Learning

Polish Academy of Sciences(波兰科学院);Institute of Computer Science(计算机科学研究所);Jagiellonian University(雅盖隆大学)

Paweł Morawiecki, Andrii Krutsylo, Maciej Wołczyk, Marek Śmieja

摘要:

持续学习旨在将机器学习带入一个更现实的场景,在这个场景中,任务是按顺序学习的,i.i.d.假设没有被保留。虽然这种设置对于生物系统来说是自然的,但对于人工神经网络等机器学习模型来说,它被证明是非常困难的。为了减少这种性能差距,我们调查的问题是否生物启发赫布学习是有用的,以应对不断的挑战。特别是,我们强调了一个现实的,往往被忽视的无监督设置,学习者必须建立表示没有任何监督。通过将稀疏神经网络与Hebbian学习原理相结合,我们建立了一个简单而有效的替代方案(HebbCL),以通过梯度下降训练的典型神经网络模型。由于Hebbian学习,网络具有易于解释的权重,这在安全或医疗保健等关键应用中可能是必不可少的。我们证明了HebbCL在应用于MNIST和Omniglot数据集的无监督学习设置中的有效性。我们还适应算法的监督的情况下,在类增量学习,并取得了可喜的成果。

英文摘要:

Continual Learning aims to bring machine learning into a more realistic scenario, where tasks are learned sequentially and the i.i.d. assumption is not preserved. Although this setting is natural for biological systems, it proves very difficult for machine learning models such as artificial neural networks. To reduce this performance gap, we investigate the question whether biologically inspired Hebbian learning is useful for tackling continual challenges. In particular, we highlight a realistic and often overlooked unsupervised setting, where the learner has to build representations without any supervision. By combining sparse neural networks with Hebbian learning principle, we build a simple yet effective alternative (HebbCL) to typical neural network models trained via the gradient descent. Due to Hebbian learning, the network have easily interpretable weights, which might be essential in critical application such as security or healthcare. We demonstrate the efficacy of HebbCL in an unsupervised learning setting applied to MNIST and Omniglot datasets. We also adapt the algorithm to the supervised scenario and obtain promising results in the class-incremental learning.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑