arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.05080cs.LGstat.ML
收藏

基于多视角任务条件神经网络的持续学习

Continual Learning Using Multi-view Task Conditional Neural Networks

University of Surrey(萨里大学);UK Dementia Research Institute(英国痴呆症研究所);Imperial College London(帝国理工学院);Adobe(奥多比公司)

Honglin Li, Payam Barnaghi, Shirin Enshaeifar, Frieder Ganz

摘要:

传统的深度学习模型在顺序学习多个任务方面的能力有限。在持续学习中,忘记先前学习的任务的问题被称为灾难性遗忘或干扰。当输入数据或学习目标发生变化时,连续模型将学习并适应新的状态。然而,模型不会记住或识别对先前状态的任何重新访问。这会导致性能降低和重新训练曲线,以处理数据或目标中的周期性或不规则重复发生的变化。在持续学习模型中,目标或数据的变化被称为新任务。大多数持续学习方法具有任务已知设置,其中任务身份对于学习模型是预先已知的。提出了一种不需要预先知道重复任务的多视图任务条件神经网络(Mv-TCNN)。我们使用MNIST,CIFAR 10,CIFAR 100标准数据集以及我们在远程医疗监测研究中收集的真实数据集(即TIHM数据集)评估我们的模型。该模型在持续学习和适应未预先定义的新任务方面优于最先进的解决方案。

英文摘要:

Conventional deep learning models have limited capacity in learning multiple tasks sequentially. The issue of forgetting the previously learned tasks in continual learning is known as catastrophic forgetting or interference. When the input data or the goal of learning change, a continual model will learn and adapt to the new status. However, the model will not remember or recognise any revisits to the previous states. This causes performance reduction and re-training curves in dealing with periodic or irregularly reoccurring changes in the data or goals. The changes in goals or data are referred to as new tasks in a continual learning model. Most of the continual learning methods have a task-known setup in which the task identities are known in advance to the learning model. We propose Multi-view Task Conditional Neural Networks (Mv-TCNN) that does not require to known the reoccurring tasks in advance. We evaluate our model on standard datasets using MNIST, CIFAR10, CIFAR100, and also a real-world dataset that we have collected in a remote healthcare monitoring study (i.e. TIHM dataset). The proposed model outperforms the state-of-the-art solutions in continual learning and adapting to new tasks that are not defined in advance.

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。

↑