We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Continual Learning Augmented Investment Decisions

Abstract: Investment decisions can benefit from incorporating an accumulated knowledge of the past to drive future decision making. We introduce Continual Learning Augmentation (CLA) which is based on an explicit memory structure and a feed forward neural network (FFNN) base model and used to drive long term financial investment decisions. We demonstrate that our approach improves accuracy in investment decision making while memory is addressed in an explainable way. Our approach introduces novel remember cues, consisting of empirically learned change points in the absolute error series of the FFNN. Memory recall is also novel, with contextual similarity assessed over time by sampling distances using dynamic time warping (DTW). We demonstrate the benefits of our approach by using it in an expected return forecasting task to drive investment decisions. In an investment simulation in a broad international equity universe between 2003-2017, our approach significantly outperforms FFNN base models. We also illustrate how CLA's memory addressing works in practice, using a worked example to demonstrate the explainability of our approach.
Comments: NeurIPS 2018 Workshop on Challenges and Opportunities for AI in Financial Services: the Impact of Fairness, Explainability, Accuracy, and Privacy, Montreal, Canada. This is a non-archival publication - the authors may submit revisions and extensions of this paper to other publication venues
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computational Finance (q-fin.CP); Portfolio Management (q-fin.PM); Trading and Market Microstructure (q-fin.TR); Machine Learning (stat.ML)
Cite as: arXiv:1812.02340 [cs.LG]
  (or arXiv:1812.02340v4 [cs.LG] for this version)

Submission history

From: Daniel Philps [view email]
[v1] Thu, 6 Dec 2018 04:26:25 GMT (502kb,D)
[v2] Sat, 8 Dec 2018 17:26:47 GMT (502kb,D)
[v3] Fri, 14 Dec 2018 10:19:57 GMT (500kb,D)
[v4] Fri, 25 Jan 2019 09:51:52 GMT (502kb,D)

Link back to: arXiv, form interface, contact.