Improving Forecasts of Suicide Attempts for Patients with Little Data
This work addresses the problem of improving suicide attempt forecasts for patients with limited data, which is an incremental advancement in mental health prediction.
The paper tackled the challenge of predicting suicide attempts from Ecological Momentary Assessment data by addressing poor performance of single models and overfitting in individualized models, introducing Latent Similarity Gaussian Processes to leverage patient similarity, with preliminary results showing it outperforms most baselines without kernel-design.
Ecological Momentary Assessment provides real-time data on suicidal thoughts and behaviors, but predicting suicide attempts remains challenging due to their rarity and patient heterogeneity. We show that single models fit to all patients perform poorly, while individualized models improve performance but still overfit to patients with limited data. To address this, we introduce Latent Similarity Gaussian Processes (LSGPs) to capture patient heterogeneity, enabling those with little data to leverage similar patients' trends. Preliminary results show promise: even without kernel-design, we outperform all but one baseline while offering a new understanding of patient similarity.