Reviews

"Irene Aldridge and Marco Avellaneda are articulate enthusiasts for Big Data Finance. They have a deep knowledge of neural networks, artificial intelligence, machine learning, and many other tools―and they are excited to share their skills. Each chapter of this wonderful book entices the reader with a broad overview, and then shows how these new concepts can be applied in financial markets. The authors are Big Data visionaries whose book belongs on your desk, not on your bookshelf." ―Elroy Dimson, Professor of Finance, Cambridge Judge Business School

"A timely, engaging, satisfying read told in a clear and lively style that wins access to a host of complex ideas. Big Data Science in Finance reaches for a broader audience than the usual subject-matter experts―and succeeds." ―Bruce Ells, VP and Director, Infrastructure Investments, TD Greystone Asset Management

"Asset managers and hedge funds are acutely aware that delivering alpha is becoming simultaneously more important and difficult. Given this background, Big Data and machine learning have become essential sources of new differentiating alpha. This much needed timely text on Big Data in finance is a refreshingly hands-on introduction to this essential subject matter that should advance the understanding of these methods and their application in modern portfolio management." ―Bernd Wuebben, Global Head, Fixed Income Quantitative Research and Systematic Investing, AllianceBernstein

"WOW! My first glance reminds me of the tried and true approach―provide theoretical background, then show implementable examples. I am actually thinking of using the book for a 'Data in Finance' offering I am working on." ―John Paul Broussard, Professor of Finance, Rutgers University and Estonian Business School

"Two of the most important figures in AI Finance have come out with a must-read Tour de Force! Soon to be a stable textbook in all of our top MBA programs." ―Jim Kyung-Soo Liew, Professor, Johns Hopkins Carey Business School

Table of Contents

For a more detailed outline, please see lesson plans

1. Why Big Data?
The changes in market structure, technology and mathematical innovation driving the trend.
2. Neural Networks in Finance
How Neural Networks work, and how to build a profitable neural network from scratch. Applications: daily price data.
3. Supervised Learning
Supervised Learning Models, step-by-step. Applications: foreign exchange, equities, fixed income pricing.
• Ridge Regression, LASSO, Elastic Nets
• K Nearest Neighbors (K-NN)
• Decision Trees, Random Decision Forests and Extra Trees
• Support Vector Machines (SVMs)
4. Modeling Human Behavior with Semi-Supervised Learning
Modeling expert researchers' decision-making process to enhance your current research framework. Applications: predicting equity ratings.
• Performance Evaluation via Cross-Validation
• Generative Models
• Discriminative Models
• Graph-based Models
5. Letting the Data Speak with Unsupervised Learning
How the artificial intelligence unsupervised learning models work, step-by-step, and why they are the future of Finance.
• Dimensionality Reduction in Finance
• Dimensionality Reduction with Unsupervised Learning
• Singular Value Decomposition
• Deconstructing Financial Returns
• Singular Vectors as Portfolio Weights
• Principal Component Regression
• Key Big Data Tools: SVD and PCA in Detail
6. Big Data Factor Models
POET and other optimal factor models. Applications: Optimal portfolio management.
• Optimal Factorization
• Eigenportfolios
• Factor Discovery
• Approximate Factor Models, Unknown Factors (POET), Instrumented PCA, The Three Pass Model, Risk-Premium PCA, Nonlinear Factorization, Projected PCA
7. Data as a Signal vs. Noise
How to determine the optimal number of factors and how to treat missing observations with the power of Big Data.
• Random data show in Eigenvalue distribution
• What's in a data bag?
• The Marcenko-Pastur Theorem
• Spike Model
• Dealing with Highly Correlated Data
• The Karhunen-Loeve Transform
• Data Imputation
• Missing Eigenvalues
• The Tracy-Widom Distribution
• Id Missing Streaming Data (the Johnson-Lindenstrauss Lemma)
8. Applications: Unsupervised Learning in Option Pricing and Stochastic Modeling
Volatility surface modeling and predicting credit rating migration.
9. Data Clustering
Applications: Optimal Portfolio Management for Cryptocurrencies and Commodities.

Lesson Plans

Download free lesson plans:

Teaching slides

Download editable slides for the course you teach.

• 1 Intro
• 2 NN
• 3 SL
• 4 SSL
• 5 UL Intro
• 6 UL Factors
• 7 UL Noise
• 8 UL Apps
• 9 Clusters

Apps

Enjoy Big Data Science in Finance hands-on

Media

Contact

We love our fans!