Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time series
Abstract Accurate time series forecasts are crucial for various applications, such as traffic management, electricity consumption, and healthcare. However, limitations in models and data quality can significantly impact forecasts’ accuracy. One common issue with data quality is the absence of data p...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-02-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-86382-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850067021582041088 |
|---|---|
| author | Saad Noufel Nadir Maaroufi Mehdi Najib Mohamed Bakhouya |
| author_facet | Saad Noufel Nadir Maaroufi Mehdi Najib Mohamed Bakhouya |
| author_sort | Saad Noufel |
| collection | DOAJ |
| description | Abstract Accurate time series forecasts are crucial for various applications, such as traffic management, electricity consumption, and healthcare. However, limitations in models and data quality can significantly impact forecasts’ accuracy. One common issue with data quality is the absence of data points, referred to as missing data values. It is often caused by sensor malfunctions, equipment failures, or human errors. This paper proposes Hinge-FM2I, a novel method for handling missing data values in univariate time series data. Hinge-FM2I builds upon the strengths of the Forecasting Method by Image Inpainting (FM2I). FM2I has proven effective, but selecting the most accurate forecasts remains a challenge. To overcome this issue, we proposed a selection algorithm. Inspired by door hinges, Hinge-FM2I drops a data point either before or after the gap (left/right-hinge), then uses FM2I for imputation. In fact, it selects the imputed gap based on the lowest error of the dropped data point. Hinge-FM2I was evaluated on a comprehensive sample composed of 1356 time series. These latter are extracted from the M3 competition benchmark dataset, with missing value rates ranging from 3.57 to 28.57%. Experimental results demonstrate that Hinge-FM2I significantly outperforms established methods such as linear/spline interpolation, K-Nearest Neighbors, and ARIMA. Notably, Hinge-FM2I achieves an average Symmetric Mean Absolute Percentage Error score of 5.6% for small gaps and up to 10% for larger ones. These findings highlight the effectiveness of Hinge-FM2I as a promising new method for addressing missing values in univariate time series data. |
| format | Article |
| id | doaj-art-ea8999ae1ce8416db59eae62bb9732bf |
| institution | DOAJ |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-ea8999ae1ce8416db59eae62bb9732bf2025-08-20T02:48:30ZengNature PortfolioScientific Reports2045-23222025-02-0115111310.1038/s41598-025-86382-4Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time seriesSaad Noufel0Nadir Maaroufi1Mehdi Najib2Mohamed Bakhouya3TICLab and LERMA Lab, College of Engineering and Architecture, International University of RabatTICLab and LERMA Lab, College of Engineering and Architecture, International University of RabatTICLab and LERMA Lab, College of Engineering and Architecture, International University of RabatTICLab and LERMA Lab, College of Engineering and Architecture, International University of RabatAbstract Accurate time series forecasts are crucial for various applications, such as traffic management, electricity consumption, and healthcare. However, limitations in models and data quality can significantly impact forecasts’ accuracy. One common issue with data quality is the absence of data points, referred to as missing data values. It is often caused by sensor malfunctions, equipment failures, or human errors. This paper proposes Hinge-FM2I, a novel method for handling missing data values in univariate time series data. Hinge-FM2I builds upon the strengths of the Forecasting Method by Image Inpainting (FM2I). FM2I has proven effective, but selecting the most accurate forecasts remains a challenge. To overcome this issue, we proposed a selection algorithm. Inspired by door hinges, Hinge-FM2I drops a data point either before or after the gap (left/right-hinge), then uses FM2I for imputation. In fact, it selects the imputed gap based on the lowest error of the dropped data point. Hinge-FM2I was evaluated on a comprehensive sample composed of 1356 time series. These latter are extracted from the M3 competition benchmark dataset, with missing value rates ranging from 3.57 to 28.57%. Experimental results demonstrate that Hinge-FM2I significantly outperforms established methods such as linear/spline interpolation, K-Nearest Neighbors, and ARIMA. Notably, Hinge-FM2I achieves an average Symmetric Mean Absolute Percentage Error score of 5.6% for small gaps and up to 10% for larger ones. These findings highlight the effectiveness of Hinge-FM2I as a promising new method for addressing missing values in univariate time series data.https://doi.org/10.1038/s41598-025-86382-4Univariate time seriesMissing data imputationInterpolationImage inpainting |
| spellingShingle | Saad Noufel Nadir Maaroufi Mehdi Najib Mohamed Bakhouya Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time series Scientific Reports Univariate time series Missing data imputation Interpolation Image inpainting |
| title | Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time series |
| title_full | Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time series |
| title_fullStr | Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time series |
| title_full_unstemmed | Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time series |
| title_short | Hinge-FM2I: an approach using image inpainting for interpolating missing data in univariate time series |
| title_sort | hinge fm2i an approach using image inpainting for interpolating missing data in univariate time series |
| topic | Univariate time series Missing data imputation Interpolation Image inpainting |
| url | https://doi.org/10.1038/s41598-025-86382-4 |
| work_keys_str_mv | AT saadnoufel hingefm2ianapproachusingimageinpaintingforinterpolatingmissingdatainunivariatetimeseries AT nadirmaaroufi hingefm2ianapproachusingimageinpaintingforinterpolatingmissingdatainunivariatetimeseries AT mehdinajib hingefm2ianapproachusingimageinpaintingforinterpolatingmissingdatainunivariatetimeseries AT mohamedbakhouya hingefm2ianapproachusingimageinpaintingforinterpolatingmissingdatainunivariatetimeseries |