Value-at-risk (VaR) is an important risk measure now widely used by financial institutions and regulators to quantify market risk and compute regulatory capital charge. The performance of VaR model can be examined by back-testing. Based on back-testing information, this paper develops a Machine...