Using K-means Clustering for Out-of-Sample Predictions of Memory Retention

AbstractIn applied settings, computational models of memory have proven useful in making principled performance predictions. Specifically, historical data are used to derive model parameters in order to enable out-of-sample predictions. Parameters are typically fit to meaningful subsets of data. However, labels that demarcate what constitutes a “meaningful” subset are not always available. Here, we utilize a data-driven method to cluster past performance into subsets possessing statistical similarities. We contrast predictions from cluster-specific model parameters with predictions based on subsets that are artifacts of the experimental design. We show that cluster-based predictions are at least as accurate as the chosen baselines and highlight additional advantages of the data-driven approach.


Return to previous page