Deep Learning in Finance

By J. B. Heaton, Nicholas G. Polson, Jan Hendrik Witte. 2016

216 citations recorded by the source corpus.

We explore the use of deep learning hierarchical models for problems in nancial prediction and classication. Financial prediction problems { such as those presented in designing and pricing securities, constructing portfolios, and risk management { often involve large data sets with complex data interactions that currently are dicult or impossible to specify in a full economic model. Applying deep learning methods to these problems can produce more useful results than standard methods in nance. In particular, deep learning can detect and exploit interactions in the data that are, at least currently, invisible to any existing nancial economic theory.

Subfields: Economics and Econometrics, Finance, Management Science and Operations Research, Economics & Operations Research

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Economics and Econometrics

In Finance

In Management Science and Operations Research

In Economics & Operations Research

Citations in the corpus

Explainable machine learning in deployment (2019, background): “Machine learning (ML) systems are being increasingly embedded into many aspects of daily life, such as healthcare [18], finance [27], and social media [6].”
Decision support from financial disclosures with deep neural networks and transfer learning (2017, methodology): “Similarly, the LSTM is applied to predict future stock returns and generates a portfolio of stocks that can yield higher returns than the top 10 stocks in the S&P 500 [34].”
Empirical Asset Pricing via Machine Learning (2018, background): “Heaton et al. (2016) develop a neural network for portfolio selection.”
Deep Portfolio Theory (2016, methodology): “This builds on deep learning methods in ﬁnance, as introduced by Heaton, Polson, and Witte (2016).”
Plant Disease Detection and Classification by Deep Learning (2019, background): “However, in the second phase, state-of-the-art algorithms/architectures were developed for many applications including self-driving cars [9–11], healthcare sector [12–14], text recognition [6,15–17], earthquake predictions [18–20], marketing [21], finance [22,23], and image recognition [24–29].”
A Survey of Deep Learning: Platforms, Applications and Emerging Research Trends (2018, unknown): “[51] considered deep learning for financial prediction and classification, particularly in the application of deep models over shallow models for high-dimensionality data reduction, high-dimensionality feature extraction for risk analysis, and event analysis classification.”
Enhancing the momentum strategy through deep regression (2019, methodology): “DNN using SAE was ﬁrst conceived in Bengio et al. (2006), and has been tested for stock price forecasting in Bao et al. (2017), for constructing portfolios that replicate stock indices in Heaton et al. (2016) and for classifying hyperspectral data in Chen et al. (2014) all with great successes.”
Stock price prediction using LSTM, RNN and CNN-sliding window model (2017, methodology): “The poposed work focuses on the application of deep learning algorithms for stock price prediction [7] [8].”
ImageNet Training in Minutes (2017, background): “Moreover, many applications, such as computational finance [14], autonomous driving [3], oil and gas exploration [15], and medical imaging [11], will almost certainly require training data-sets with billions of training elements and”
Quantum computing for finance: Overview and prospects (2018, background): “[13] offers a detailed description of deep learning in finance.”
Performance evaluation and hyperparameter tuning of statistical and machine-learning models using spatial data (2018, methodology): “Statistical learning has become an important tool in the process of knowledge discovery from big data in fields as diverse as finance or geomarketing (Heaton et al., 2016; Schernthanner et al., 2017), medicine (Leung et al.”
Financial Time Series Prediction Based on Deep Learning (2017, background): “Second, deep learning is more likely to alleviate the overfitting problem while improving the accuracy of sample prediction [8], and thus have a stronger generalization ability.”