HP Trend Filtering Using Gaussian Mixture Model Weighted Heuristic

Luiza Sayfullina, Magnus Westerlund , Kaj-Mikael Björk, Hannu T. Toivonen

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review


Trends show the underlying structure of the time series data. Trend estimation is a commonly used tool for financial market movement prediction. In traditional approaches, such as Hodrick-Prescott (HP) and L1 filtering, the trend is considered as a smoothed version of the time-series, including rare significant hills that are smoothed in the same way as usual noise. The goal of this paper is to allow the estimated trend to be more complex and detailed in the intervals of significant changes while making a smooth estimate in all other parts. This will be our main criteria for trend estimation. We present a modified version of HP weighted heuristic that provides the best trend according to the abovementioned criteria. Gaussian Mixture Models (GMMs) on the preliminary estimated trend are used in the weighted HP heuristic to decrease the penalty in the objective function for turning-point intervals. We conducted a set of experiments on financial datasets and compared the results with those obtained from the standard HP filtering with weighted heuristic. The results indicate an improvement in the cycling component using our proposed criteria compared to the HP filtering approach.
Original languageEnglish
Title of host publication 2014 IEEE 26th International Conference on Tools with Artificial Intelligence
Publication date15.12.2014
ISBN (Electronic)978-1-4799-6572-4
Publication statusPublished - 15.12.2014
MoE publication typeA4 Article in conference proceedings
Event2014 IEEE 26th International Conference on Tools with Artificial Intelligence - Limassol, Cyprus
Duration: 10.11.201412.11.2014
Conference number: 26


  • 512 Business and Management


Dive into the research topics of 'HP Trend Filtering Using Gaussian Mixture Model Weighted Heuristic'. Together they form a unique fingerprint.

Cite this