| Content | In this event, we will host the second practice session on the implementation of convolutional neural network (CNN) and recurrent neural network (RNN) using Keras python package. It aims to equip the participants with the hands-on experience in supervised learning with Keras and deepen the understanding on the last lecture session, especially the application of Recurrent Reinforcement Direct Learning (RRL) in trading.Dr Hao Ni will give a presentation based on the Google Colaboratory Python demo in order to(1) show how to implement and fine tune the CNN models via the working example on the standard image classification dataset, i.e. Cifar10;(2) and reformulate the RRL model as the customized neural network model with RNN layer. Following [1] we will apply the RRL algorithm to learn optimal trading strategies on synthetic data.We expect all participants bring their laptops to the session and follow the instructions to work through the working examples and play with the datasets by themselves. After the session, the participants are expected to independently(1) implement CNN and RNN to solve classification problems using Keras;(2) write the customized neural network models using Keras;(3) and with the twist of RNN models, implement the RRL and apply it to the trading applications. It is followed by the stimulating group discussion.Reference1. Moody, J. and Saffell, M., 2001. Learning to trade via direct reinforcement. IEEE transactions on neural Networks, 12(4), pp.875-889. 2. https://colab.research.google.com/.3. https://github.com/ageron/handson-ml4. Géron, A., 2017. Hands-on machine learning with Scikit-Learn and TensorFlow: concepts, tools, and techniques to build intelligent systems. " O'Reilly Media, Inc."5. https://colab.research.google.com/Link: https://www.eventbrite.co.uk/e/meetup-for-machine-learning-in-quantitative-finance-tickets-49216667377Passowrd: MachineLearningQuantFinance1 |