GPR works well with small datasets and generates a metric of confidence of a predicted result, but it's moderately complex and the results are not easily interpretable, says Dr. James McCaffrey of ...

Modeling counterparty risk is computationally challenging because it requires the simultaneous evaluation of all trades between each counterparty under both market and credit risk. We present a ...

Gaussian process regression is a sophisticated technique that uses what is called the kernel trick to deal with complex non-linear data, and L2 regularization to avoid model overfitting where a model ...