<ul><li>Traditional neural network regression models provide point estimates, lacking predictive uncertainty.</li><li>Probabilistic neural networks (PNNs) address this by producing output distributions, allowing prediction intervals.</li><li>To improve adaptability, t-Distributed Neural Networks (TDistNNs) are proposed, generating t-distributed outputs.</li><li>TDistNNs produce narrower prediction intervals compared to Gaussian-based PNNs while maintaining proper coverage.</li></ul>

Probabilistic Neural Networks (PNNs) with t-Distributed Outputs: Adaptive Prediction Intervals Beyond Gaussian Assumptions

Discover more