MRFPY

MRFPY: Multi Resolution Forecasting in Python

MRFR: Multi Resolution Forecasting in R

Multi Resolution Time Series Forecasting (with wavelets)

This is a Python package for univariate time series forecasting. There is also a R version of it. They are similar in naming and handling.

Contents

1. Description
2. Use cases
3. Installation
4. References

Description

This github repository provides an implementation of the algorithm of the workgroup around F. Murtagh. It uses a redundant Haar wavelet transform to decompose a time series in its wavelet and the corresponding smooth approximation features. Those features are processed in linear or nonlinear methods in order to yield a one-step forecast. Multi-step forecasts are obtained recursively. Currently, only univariate time series can be forecasted. There is ongoing work for multivariate time series forecasting. Find the theoretical work from Murtagh et al. in the references.

You can create one-step forecasts with various linear and nonlinear methods using wavelet features trying out different possibilites. One-step forecasts can be created by directly accessing the methods specific function call or the abstract method "onestep". Multi-step forecasts are computed recursively and can be called with the abstract method "multistep". Evaluation studies of one specific setting can be computed with the rolling window function. A complete model selection with nested cross validation can be called with the function model_selection.

Use cases

4 use cases: Callcenter, Electricity demand, oil prices and stock values.

Installation

Execute following command in a terminal:

pip install MRFPY

References

Aussem, A., Campbell, J., and Murtagh, F. (1998) Wavelet-based feature extraction and decomposition strategies for financial forecasting, International Journal of Computational Intelligence in Finance, 6 (5-12).

Aussem, A., Campbell, J., and Murtagh, F.: Waveletbased Feature Extraction and Decomposition Strategies for Financial Forecasting. International Journal of Computational Intelligence in Finance, 6:5–12. 1998.

Benaouda, D., Murtagh, F., Starck, J.-L., and Renaud, O.: Wavelet-based Nonlinear Multiscale Decomposition Model for Electricity Load Forecasting. Neurocomputing, 70(1-3):139–154. doi:10.1016/j.neucom.2006.04.005. 2006.

Gonghui, Z., Starck, J.-L., Campbell, J., and Murtagh, F.: The Wavelet Transform for Filtering Financial Data Streams. Journal of Computational Intelligence in Finance, 7(3):18–35. 1999.

Murtagh, F., Starck, J.-L., and Renaud, O.: On Neuro-Wavelet Modeling. Decision Support Systems, 37(4):475–484. doi:10.1016/S0167-9236(03)00092-7. 2004.

Renaud, O., Starck, J.-L., and Murtagh, F.: Prediction based on a Multiscale Decomposition. International Journal of Wavelets, Multiresolution and Information Processing, 1(2):217–232. doi:10.1142/S0219691303000153. 2003.

Renaud, O., Starck, J.-L., and Murtagh, F.: Wavelet-based combined Signal Filtering and Prediction. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 35(6):1241–1251. doi:10.1109/TSMCB.2005.850182. 2005.