Fig. 1
From: CLeaR: An adaptive continual learning framework for regression tasks
Distributions of the inputs change over time. We split 10000 samples into four sections sequentially over time, each of which has 2500 samples, and visualize their two-dimensional distributions from (a) to (d). (e) illustrates the distribution of the same 10000 samples. Both axes in the sub-figures, X1 and X2, indicate the first two principal components extracted from seven-dimensional weather features