Probabilistic prediction with locally weighted jackknife predictive system
Abstract Probabilistic predictions for regression problems are more popular than point predictions and interval predictions, since they contain more information for test labels. Conformal predictive system is a recently proposed non-parametric method to do reliable probabilistic predictions, which i...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2023-04-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-023-01044-0 |
| _version_ | 1797674979184082944 |
|---|---|
| author | Di Wang Ping Wang Pingping Wang Cong Wang Zhen He Wei Zhang |
| author_facet | Di Wang Ping Wang Pingping Wang Cong Wang Zhen He Wei Zhang |
| author_sort | Di Wang |
| collection | DOAJ |
| description | Abstract Probabilistic predictions for regression problems are more popular than point predictions and interval predictions, since they contain more information for test labels. Conformal predictive system is a recently proposed non-parametric method to do reliable probabilistic predictions, which is computationally inefficient due to its learning process. To build faster conformal predictive system and make full use of training data, this paper proposes the predictive system based on locally weighted jackknife prediction approach. The theoretical property of our proposed method is proved with some regularity assumptions in the asymptotic setting, which extends our earlier theoretical researches from interval predictions to probabilistic predictions. In the experimental section, our method is implemented based on our theoretical analysis and its comparison with other predictive systems is conducted using 20 public data sets. The continuous ranked probability scores of the predictive distributions and the performance of the derived prediction intervals are compared. The better performance of our proposed method is confirmed with Wilcoxon tests. The experimental results demonstrate that the predictive system we proposed is not only empirically valid, but also provides more information than the other comparison predictive systems. |
| first_indexed | 2024-03-11T22:08:02Z |
| format | Article |
| id | doaj.art-8131671b88f44fed85d20e5e09363249 |
| institution | Directory Open Access Journal |
| issn | 2199-4536 2198-6053 |
| language | English |
| last_indexed | 2024-03-11T22:08:02Z |
| publishDate | 2023-04-01 |
| publisher | Springer |
| record_format | Article |
| series | Complex & Intelligent Systems |
| spelling | doaj.art-8131671b88f44fed85d20e5e093632492023-09-24T11:35:40ZengSpringerComplex & Intelligent Systems2199-45362198-60532023-04-01955761577810.1007/s40747-023-01044-0Probabilistic prediction with locally weighted jackknife predictive systemDi Wang0Ping Wang1Pingping Wang2Cong Wang3Zhen He4Wei Zhang5School of Electrical and Information Engineering, Tianjin UniversitySchool of Electrical and Information Engineering, Tianjin UniversityQingdao Academy of Chinese Medical Science, Shandong University of Traditional Chinese MedicineSchool of Electrical and Information Engineering, Tianjin UniversityCollege of Management and Economics, Tianjin UniversityCollege of Management and Economics, Tianjin UniversityAbstract Probabilistic predictions for regression problems are more popular than point predictions and interval predictions, since they contain more information for test labels. Conformal predictive system is a recently proposed non-parametric method to do reliable probabilistic predictions, which is computationally inefficient due to its learning process. To build faster conformal predictive system and make full use of training data, this paper proposes the predictive system based on locally weighted jackknife prediction approach. The theoretical property of our proposed method is proved with some regularity assumptions in the asymptotic setting, which extends our earlier theoretical researches from interval predictions to probabilistic predictions. In the experimental section, our method is implemented based on our theoretical analysis and its comparison with other predictive systems is conducted using 20 public data sets. The continuous ranked probability scores of the predictive distributions and the performance of the derived prediction intervals are compared. The better performance of our proposed method is confirmed with Wilcoxon tests. The experimental results demonstrate that the predictive system we proposed is not only empirically valid, but also provides more information than the other comparison predictive systems.https://doi.org/10.1007/s40747-023-01044-0Probabilistic predictionPredictive systemJackknife predictionAsymptotic analysisConformal prediction |
| spellingShingle | Di Wang Ping Wang Pingping Wang Cong Wang Zhen He Wei Zhang Probabilistic prediction with locally weighted jackknife predictive system Complex & Intelligent Systems Probabilistic prediction Predictive system Jackknife prediction Asymptotic analysis Conformal prediction |
| title | Probabilistic prediction with locally weighted jackknife predictive system |
| title_full | Probabilistic prediction with locally weighted jackknife predictive system |
| title_fullStr | Probabilistic prediction with locally weighted jackknife predictive system |
| title_full_unstemmed | Probabilistic prediction with locally weighted jackknife predictive system |
| title_short | Probabilistic prediction with locally weighted jackknife predictive system |
| title_sort | probabilistic prediction with locally weighted jackknife predictive system |
| topic | Probabilistic prediction Predictive system Jackknife prediction Asymptotic analysis Conformal prediction |
| url | https://doi.org/10.1007/s40747-023-01044-0 |
| work_keys_str_mv | AT diwang probabilisticpredictionwithlocallyweightedjackknifepredictivesystem AT pingwang probabilisticpredictionwithlocallyweightedjackknifepredictivesystem AT pingpingwang probabilisticpredictionwithlocallyweightedjackknifepredictivesystem AT congwang probabilisticpredictionwithlocallyweightedjackknifepredictivesystem AT zhenhe probabilisticpredictionwithlocallyweightedjackknifepredictivesystem AT weizhang probabilisticpredictionwithlocallyweightedjackknifepredictivesystem |