Twin neural network regression is a semi-supervised regression algorithm
Sebastian J. Wetzel, Roger G. Melko, Isaac Tamblyn
Abstract
Abstract Twin neural network regression (TNNR) is trained to predict differences between the target values of two different data points rather than the targets themselves. By ensembling predicted differences between the targets of an unseen data point and all training data points, it is possible to obtain a very accurate prediction for the original regression problem. Since any loop of predicted differences should sum to zero, loops can be supplied to the training data, even if the data points themselves within loops are unlabelled. Semi-supervised training improves TNNR performance, which is already state of the art, significantly.
Topics & Concepts
RegressionArtificial neural networkArtificial intelligenceComputer scienceData pointRegression analysisTraining setPoint (geometry)Linear regressionPattern recognition (psychology)Machine learningAlgorithmMathematicsStatisticsGeometryNeural Networks and ApplicationsFace and Expression RecognitionMachine Learning and ELM