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 2007.07617cs.LGcs.CVstat.ML
收藏

SpaceNet:为持续学习腾出空间

SpaceNet: Make Free Space For Continual Learning

Eindhoven University of Technology(埃因霍温理工大学);University of Twente(特文特大学)

Ghada Sokar, Decebal Constantin Mocanu, Mykola Pechenizkiy

摘要:

持续学习(CL)范式旨在使神经网络能够以顺序的方式连续地学习任务。这种学习范式的根本挑战是,当模型针对新任务进行优化时,会发生灾难性遗忘先前学习的任务,特别是当它们的数据不可访问时。当前基于架构的方法旨在缓解灾难性遗忘问题,但以扩展模型的容量为代价。基于正则化的方法保持固定的模型容量;然而,以前的研究表明,当任务身份在推理过程中不可用时(例如类增量学习场景),这些方法的性能会大幅下降。在这项工作中,我们提出了一种新的基于架构的方法称为SpaceNet类增量学习的情况下,我们利用可用的固定容量的模型智能。SpaceNet以自适应的方式从头开始训练稀疏的深度神经网络,将每个任务的稀疏连接压缩到紧凑的神经元数量中。稀疏连接的自适应训练导致减少任务之间干扰的稀疏表示。实验结果表明,我们提出的方法对灾难性遗忘旧任务的鲁棒性和效率的空间网络在利用模型的可用容量,留下更多的任务学习的空间。特别是,当SpaceNet在CL的著名基准测试上进行测试时:分裂MNIST,分裂Fashion-MNIST和CIFAR-10/100,它的性能优于基于正则化的方法。此外,它实现了更好的性能比基于架构的方法没有模型扩展,并取得了类似的结果与排练为基础的方法,同时提供了一个巨大的内存减少。

英文摘要:

The continual learning (CL) paradigm aims to enable neural networks to learn tasks continually in a sequential fashion. The fundamental challenge in this learning paradigm is catastrophic forgetting previously learned tasks when the model is optimized for a new task, especially when their data is not accessible. Current architectural-based methods aim at alleviating the catastrophic forgetting problem but at the expense of expanding the capacity of the model. Regularization-based methods maintain a fixed model capacity; however, previous studies showed the huge performance degradation of these methods when the task identity is not available during inference (e.g. class incremental learning scenario). In this work, we propose a novel architectural-based method referred as SpaceNet for class incremental learning scenario where we utilize the available fixed capacity of the model intelligently. SpaceNet trains sparse deep neural networks from scratch in an adaptive way that compresses the sparse connections of each task in a compact number of neurons. The adaptive training of the sparse connections results in sparse representations that reduce the interference between the tasks. Experimental results show the robustness of our proposed method against catastrophic forgetting old tasks and the efficiency of SpaceNet in utilizing the available capacity of the model, leaving space for more tasks to be learned. In particular, when SpaceNet is tested on the well-known benchmarks for CL: split MNIST, split Fashion-MNIST, and CIFAR-10/100, it outperforms regularization-based methods by a big performance gap. Moreover, it achieves better performance than architectural-based methods without model expansion and achieved comparable results with rehearsal-based methods, while offering a huge memory reduction.

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。

↑