Take your quantitative strategies to the next level by exploring nine examples that make use of cutting-edge deep learning technologies, including CNNs, LSTMs, GANs, reinforcement learning, and CapsNets

Key Features

Implement deep learning techniques and algorithms to build financial models
Apply modern AI techniques in quantitative market modeling and investment decision making
Leverage Python libraries for rapid development and prototyping

Book Description

Quantitative methods are the vanguard of the investment management industry. This book shows how to enhance trading strategies and investments in financial markets using deep learning algorithms.

This book is an excellent reference to understand how deep learning models can be leveraged to capture insights from financial data. You will implement deep learning models using Python libraries such as TensorFlow and Keras. You will learn various deep learning algorithms to build models for understanding financial market dynamics and exploiting them in a systematic manner. This book takes a pragmatic approach to address various aspects of asset management. The information content in non-structured data like news flow is crystalized using BLSTM. Autoencoders for efficient index replication is discussed in detail. You will use CNN to develop a trading signal with simple technical indicators, and improvements offered by more complex techniques such as CapsNets. Volatility is given due emphasis by demonstrating the superiority of forecasts employing LSTM, and Monte Carlo simulations using GAN for value at risk computations. These are then brought together by implementing deep reinforcement learning for automated trading.

This book will serve as a continuing reference for implementing deep learning models to build investment strategies.

What you will learn

Implement quantitative financial models using the various building blocks of a deep neural network
Build, train, and optimize deep networks from scratch
Use LSTMs to process data sequences such as time series and news feeds
Implement convolutional neural networks (CNNs), CapsNets, and other models to create trading strategies
Adapt popular neural networks for pattern recognition in finance using transfer learning
Automate investment decisions by using reinforcement learning
Discover how a risk model can be constructed using D-GAN

Who this book is for

If you're a finance or investment professional who wants to lead the development of quantitative strategies, this book is for you. With this practical guide, you'll be able to use deep learning methods for building financial models and incorporating them in your investment process. Anyone who wants to enter the fascinating domain of quantitative finance using the power of deep learning algorithms and techniques will also find this book useful. Basic knowledge of machine learning and Python programming is required. Luigi Troiano, Ph.D., is an Associate Professor of Artificial Intelligence, Data Science, and Machine Learning at the University of Salerno (Italy), Dept. of Management and Innovation Systems. He is a coordinator of Computational and Intelligent System Engineering Lab at the University of Sannio and an NVIDIA Deep Learning Institute University Ambassador. He is also the chairman of the ISO/JTC 1/SC 42, AI and Big Data, Italian section. Arjun Bhandari is Chief Investment Officer of a family office. His previous positions have been Head of Quantitative Strategies at ADIA ( largest sovereign wealth fund in the middle east ) and APG Investments ( largest pension plan in Europe ). He has been deploying quantitative techniques for multi-asset class investments for over 20 years, bringing this experience to bear on his most recent focus on machine learning applied to fund management. Elena Mejuto Villa, Ph.D., is a data scientist in the Advanced Analytics team for Technology Services Consulting in a multinational firm in Milan. She completed her Master's Degree in Telecommunication Engineering at the University of Oviedo (Spain), and she received her Ph.D. in Information Technologies for Engineering from the University of Sannio (Italy). During her Ph.D., she researched the application of machine learning and signal processing techniques to time-varying signals/data in the fields of finance and gravitational wave data analysis.