Cypress Point


     Cypress Point Technologies, LLC
            Home of HedgeTools ®



.

Research in Deep Neural Networks

Research is being conduction using Deep Neural Networks to find momentum patterns in daily trade data.  We are using the following neural network systems:

  1. Tensorflow (http://www.tensorflow.org)

  2. H2O (http://www.h2o.ai)

  3. Applying Deep Learning to Enhance Momentum Trading Strategies in Stocks (PDF)

  4. Deep Learning for Multivariate Financial Time Series (PDF)


Data Smoothing

The HedgeTools Momentum Model is an implementation of a non-linear regression algorithm. It is based on current research that all began with the following seminal research papers and books:

  1. Hutchinson , M.F., Hoog, F.R.: Smoothing Noisy Data with Spline Functions. Numerische Mathematik 47, 99-106 1985 (PDF)

  2. Craven, P., Wahba, G.: Smoothing Noisy Data with Spline Functions. Numerische Mathematik 31, 377-403, 1979  (PDF)

  3. Eubank, Randall L., Approximate Regression Models and Splines. Tech. Report 180, Office of Naval Research. 1983  (PDF)

  4. Marsh, C., Cormier, D.: Spline Regression Models. Quantitative Applications in the Social Sciences. SAGE University Monograph, vol 137, 69  pages. (Amazon)

  5. Eubank, R., Nonparametric Regression and Spline Smoothing, Second Edition (Statistics: a Series of Textbooks and Monogrphs) 1999 (Amazon)

  6. Wahba, G.: Spline Models for Observational Data (CBMS-NSF Regional Conference Series in Applied Mathematics). 1990 (Amazon)