Dynamical Effect on Tick-Wise Price Predictions

  • Authors:

  • Mieko Tanaka-Yamawaki;Keita Awaji

  • Affiliations:

  • Department of Information and Knowledge Engineering, Tottori University, Tottori, Japan 680-8550;Fujitsu Ten, Inc., Kobe, Hyogo, Japan 652-8510

  • Venue:
  • KES '08 Proceedings of the 12th international conference on Knowledge-Based Intelligent Information and Engineering Systems, Part II
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

We attempt to incorporate dynamical effect to the tick-wise price prediction, in order to improve performance of the autonomous price prediction generator we construct. Our assumption is that such dynamical effect is carried by local parameters including derivatives of the price (velocities and accelerations) in addition to the price itself, and two dimensionless parameters constructed by using the derivatives. For this purpose, we add a new procedure to the prediction generator that computes derivatives of the data from each segment of the price time series and label that segment by those dynamical parameters. We show in this paper that this dynamical version of the price generator indeed performs better compared to the old version.