模块化持续学习的概率框架
A Probabilistic Framework for Modular Continual Learning
AI总结：
PICLE提出用概率模型低成本评估模块组合适应度，首次在模块化持续学习中同时实现感知、少样本和潜在迁移，并在长问题序列上超越现有方法。
摘要：
在持续学习（CL）中，针对每个问题使用不同模块组合的模块化方法是一个有前景的方向。然而，在庞大且离散的模块组合空间中进行搜索具有挑战性，尤其是因为评估一个组合的性能需要一轮神经网络训练。我们通过一个模块化持续学习框架PICLE来应对这一挑战，该框架使用概率模型以低成本计算每个组合的适应度，从而使PICLE能够同时实现感知迁移、少样本迁移和潜在迁移。该模型将关于良好模块组合的先验知识与数据集特定信息相结合。我们使用两个基准测试套件对PICLE进行评估，这些套件旨在评估持续学习技术的不同需求。与广泛的方法进行比较后，我们表明PICLE是第一个在长问题序列上实现感知迁移、少样本迁移和潜在迁移，同时能够良好扩展到大规模搜索空间的模块化持续学习算法，超越了先前最先进的模块化持续学习方法。
英文摘要：
Modular approaches that use a different composition of modules for each problem are a promising direction in continual learning (CL). However, searching through the large, discrete space of module compositions is challenging, especially because evaluating a composition's performance requires a round of neural network training. We address this challenge through a modular CL framework, PICLE, that uses a probabilistic model to cheaply compute the fitness of each composition, allowing PICLE to achieve both perceptual, few-shot and latent transfer. The model combines prior knowledge about good module compositions with dataset-specific information. We evaluate PICLE using two benchmark suites designed to assess different desiderata of CL techniques. Comparing to a wide range of approaches, we show that PICLE is the first modular CL algorithm to achieve perceptual, few-shot and latent transfer while scaling well to large search spaces, outperforming previous state-of-the-art modular CL approaches on long problem sequences.
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。
