Peter Bloem<p>Now out in <a href="https://sigmoid.social/tags/TMLR" class="mention hashtag" rel="tag">#<span>TMLR</span></a>:</p><p> 🍇 GRAPES: Learning to Sample Graphs for Scalable Graph Neural Networks 🍇</p><p>There&#39;s lots of work on sampling subgraphs for GNNs, but relatively little on making this sampling process _adaptive_. That is, learning to select the data from the graph that is relevant for your task.</p><p>We introduce an RL-based and a GFLowNet-based sampler and show that the approach performs well on heterophilic graphs. </p><p><a href="https://openreview.net/forum?id=QI0l842vSq" target="_blank" rel="nofollow noopener" translate="no"><span class="invisible">https://</span><span class="ellipsis">openreview.net/forum?id=QI0l84</span><span class="invisible">2vSq</span></a></p><p><a href="https://sigmoid.social/tags/machinelearning" class="mention hashtag" rel="tag">#<span>machinelearning</span></a> <a href="https://sigmoid.social/tags/graphs" class="mention hashtag" rel="tag">#<span>graphs</span></a> <a href="https://sigmoid.social/tags/graph_learning" class="mention hashtag" rel="tag">#<span>graph_learning</span></a> <a href="https://sigmoid.social/tags/paper" class="mention hashtag" rel="tag">#<span>paper</span></a></p>