The authors in the working paper were: Dr. Shatha Qamhieh Head of the Finance and Banking Department and Dr. Haya Sammaneh, Dr. Emad Natsheh and Dr. Manar Qamhieh from the Department of Computer Engineering.

The working paper uses a two-step classical-neural modelling to improve the accuracy of Bitcoin option prices forecasting by augmenting the classical option pricing models with a multilayer neural network. However, the main focus is not on the classical models performance themselves, nor on proving that neural networks are better in terms of option pricing prediction, but improving the prediction produced by the neural network through the improvement of its input features, structure, activation functions and training algorithm. 

For more information, please visit the link:

http://blockchain.cs.ucl.ac.uk/ai-financial-automation-market-risk/