https://nalar.dev/tags/consistency-regularization/ Recent content in Consistency Regularization on Nalar Hugo en-us Fri, 04 Sep 2026 00:00:00 +0700 https://nalar.dev/learn-from-unlabeled-data-with-mean-teacher/ Fri, 04 Sep 2026 00:00:00 +0700 https://nalar.dev/learn-from-unlabeled-data-with-mean-teacher/ <p>Many machine learning projects have far more raw examples than labeled ones. A team may have millions of images, audio clips, or sensor readings, but only a small subset has been reviewed by people. Standard supervised training ignores the unlabeled remainder because it has no target labels to compare with the model&rsquo;s predictions.</p> <p><strong>Mean Teacher</strong> provides a way to use those unlabeled examples without pretending that their unknown labels are known. It trains a student model to make predictions that stay consistent with a more slowly changing teacher model. The teacher is not a separately trained expert: its parameters are an exponential moving average of the student&rsquo;s parameters.</p>