Second, we t = rt+1+ t+1x>t+1w x>t w , TD-errorforw (see (2)). This t is tobelarger t.

Neural Information Processing Systems http://nips.cc/

We introduce a new neural network-based continual learning algorithm, dubbed as Uncertainty-regularized Continual Learning (UCL), which builds on traditional Bayesian online learning framework with variational inference. We focus on two significant drawbacks of the recently proposed regularization-based methods: a) considerable additional memory cost for determining the per-weight regularization strengths and b) the absence of gracefully forgetting scheme, which can prevent performance degradation in learning new tasks. In this paper, we show UCL can solve these two problems by introducing a fresh interpretation on the Kullback-Leibler (KL) divergence term of the variational lower bound for Gaussian mean-field approximation. Based on the interpretation, we propose the notion of node-wise uncertainty, which drastically reduces the number of additional parameters for implementing per-weight regularization. Moreover, we devise two additional regularization terms that enforce \emph{stability} by freezing important parameters for past tasks and allow \emph{plasticity} by controlling the actively learning parameters for a new task. Through extensive experiments, we show UCL convincingly outperforms most of recent state-of-the-art baselines not only on popular supervised learning benchmarks, but also on challenging lifelong reinforcement learning tasks. The source code of our algorithm is available at https://github.com/csm9493/UCL.

To achieve such a goal, directed exploration strategies are key.

The recent state-of-the-art regularization-based methods typically implement the per-parameter regularization parameters based on several different principles inferring the importance of each parameter for given tasks; e.g., diagonal Fisher information matrix for EWC [

This paper proposed uncertainty-regularized continue learning (UCL) to address the challenge of catastrophe forgetting of neural networks. In detail, the method improves over variational continual learning (VCL) by modifying the KL regularizer in mean-field Gaussian prior/posterior setting. The approach is mainly justified by intuition explanation rather than theoretical/mathematical arguments. Experiments are performed on supervised continual learning benchmarks (split and permuted MNIST), and the method shows dominating performance over previous baselines (VCL, SI, EWC, HAT). Reviewers include experts in continual learning.

We introduce a new neural network-based continual learning algorithm, dubbed as Uncertainty-regularized Continual Learning (UCL), which builds on traditional Bayesian online learning framework with variational inference. We focus on two significant drawbacks of the recently proposed regularization-based methods: a) considerable additional memory cost for determining the per-weight regularization strengths and b) the absence of gracefully forgetting scheme, which can prevent performance degradation in learning new tasks. In this paper, we show UCL can solve these two problems by introducing a fresh interpretation on the Kullback-Leibler (KL) divergence term of the variational lower bound for Gaussian mean-field approximation. Based on the interpretation, we propose the notion of node-wise uncertainty, which drastically reduces the number of additional parameters for implementing per-weight regularization. Moreover, we devise two additional regularization terms that enforce \emph{stability} by freezing important parameters for past tasks and allow \emph{plasticity} by controlling the actively learning parameters for a new task.

The principle that governs unsupervised multilingual learning (UCL) in jointly trained language models (mBERT as a popular example) is still being debated. Many find it surprising that one can achieve UCL with multiple monolingual corpora. In this work, we anchor UCL in the context of language decipherment and show that the joint training methodology is a decipherment process pivotal for UCL. In a controlled setting, we investigate the effect of different decipherment settings on the multilingual learning performance and consolidate the existing opinions on the contributing factors to multilinguality. From an information-theoretic perspective we draw a limit to the UCL performance and demonstrate the importance of token alignment in challenging decipherment settings caused by differences in the data domain, language order and tokenization granularity. Lastly, we apply lexical alignment to mBERT and investigate the contribution of aligning different lexicon groups to downstream performance.

Dr. Craig Basman discusses new life-saving technology and the variables that can predict sudden cardiac events. "Heart failure" is a catch-all term used to describe any condition in which the organ doesn't work as it's supposed to -- but one person's experience with the disease can be very different from someone else's. Researchers from the University College London (UCL) recently used machine learning -- a type of artificial intelligence -- to pinpoint five distinct types of heart failure, with the goal of predicting the prognosis for the different kinds. "We sought to improve how we classify heart failure, with the aim of better understanding the likely course of disease and communicating this to patients," said lead author Professor Amitava Banerjee from UCL in a press release announcing the study. "Currently, how the disease progresses is hard to predict for individual patients," he also said.