Machine Learning for Algorithmic Trading

Deep neural networks (DNNs) are powerful types of artificial neural networks (ANNs) that use several hidden layers. They have recently gained considerable attention in the speech transcription and image recognition community (Krizhevsky et al., 2012) for their superior predictive properties including robustness to overfitting. However their application to financial market prediction has not been previously researched, partly because of their computational complexity. This paper describes the application of DNNs to predicting financial market movement directions. A critical step in the viability of the approach in practice is the ability to effectively deploy the algorithm on general purpose high performance computing infrastructure. Using an Intel Xeon Phi co-processor with 61 cores, we describe the process for efficient implementation of the batched stochastic gradient descent algorithm and demonstrate a 11.4x speedup on the Intel Xeon Phi over a serial implementation on the Intel Xeon. Learn more here.

Using GPUs for Computational Finance

Many modeling applications in computational finance are latency sensitive. That is, the model must be run on-demand or in near real-time. Furthermore, such models are compute intensive to calibrate and run. At the same time, many-core processor accelerators are widely available to provide the computational resources required by financial applications. The purpose of this work is to investigate many of the parallel program design and implementational aspects specific to particular financial applications. The long term goal of this research is to address the 'implemention gap' between high level mathematical and statistical modeling environments and many-core processing environments.

Credit Modeling for Online Consumer Loans

Institutional investors and investment managers seek to better characterize the credit risk of online consumer loans. This article describes how to prepare the data and build a credit risk model that can be used for a number of applications including generating alpha, issuing protection and securitizing loans into bonds with the desired risk/reward profile. A simple example is used to provide insight into the modeling approach. In the next blog article, we shall describe how to price consumer loans and include pre-prepayment models. More details of this research.

A Bayesian Approach to Ranking Private Companies Based on Predictive Indicators

Private equity investors seek to identify potential investment opportunities in growth stage private companies and rank their prospects relative to a cohort of companies such as an industry sector. Growth stage private companies often have investment transaction histories from which industry specific characteristics associated with successful and failed companies may be discerned using machine learning. This research partially addresses one of the primary challenges in pursuing this approach - the sparsity of historical data on private companies is exacerbated in nascent sectors because of the relatively few number of observed exit events. More details of this research.

Enabling Larger Scale Water Resource Simulations with Linear Solvers

With Zhaojun Bai (CS, UC Davis), we investigated the effect of scaling PGMRES solvers applied to large datasets from water resource management tools developed by the California State Department of Water Resources. We found that row equilibration is important for sharpening upper bound estimates of the forward error. Sharp control of the forward error enables the residual stopping tolerance to be more efficiently choosen, avoiding excessive iterations and leading to a factor of 7x speedup against their baseline implementation. More details of this project.

Detecting Mobility Patterns in Mobile Phone Data from the Ivory Coast

By leveraging cloud-based open-source analytics infrastructure to (1) merge the D4D datasets with Geographic Information System (GIS) data and (2) apply data mining algorithms, this research project presents a number of techniques for detecting mobility patterns of anonymized Orange mobile phone customers in the Ivory Coast. More details of this research.