SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival Models
A new method called the Survival Beran-based Neural Importance Model (SurvBeNIM) is proposed. It aims to explain predictions of machine learning survival models, which are in the form of survival or cumulative hazard functions. The main idea behind SurvBeNIM is to extend the Beran estimator by incor...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10858699/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823859643168849920 |
|---|---|
| author | Lev V. Utkin Danila Y. Eremenko Andrei V. Konstantinov Vladimir A. Muliukha |
| author_facet | Lev V. Utkin Danila Y. Eremenko Andrei V. Konstantinov Vladimir A. Muliukha |
| author_sort | Lev V. Utkin |
| collection | DOAJ |
| description | A new method called the Survival Beran-based Neural Importance Model (SurvBeNIM) is proposed. It aims to explain predictions of machine learning survival models, which are in the form of survival or cumulative hazard functions. The main idea behind SurvBeNIM is to extend the Beran estimator by incorporating the importance functions into its kernels and by implementing these importance functions as a set of neural networks which are jointly trained in an end-to-end manner. Two strategies of using and training the whole neural network implementing SurvBeNIM are proposed. The first one explains a single instance, and the neural network is trained for each explained instance. According to the second strategy, the neural network only learns once on all instances from the dataset and on all generated instances. Then the neural network is used to explain any instance in a dataset domain. Various numerical experiments compare the method with different existing explanation methods developed for survival analysis, in particular, for models SurvBex, SurvNAM, and SurvLIME. Results of experiments with synthetic and real data show that the proposed method outperforms existing explanation methods. A code implementing the proposed method is publicly available. |
| format | Article |
| id | doaj-art-cc6024bcc3fb4e5cadcbc12a6dba93ab |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-cc6024bcc3fb4e5cadcbc12a6dba93ab2025-02-11T00:00:38ZengIEEEIEEE Access2169-35362025-01-0113241372415710.1109/ACCESS.2025.353745910858699SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival ModelsLev V. Utkin0Danila Y. Eremenko1Andrei V. Konstantinov2https://orcid.org/0000-0002-1542-6480Vladimir A. Muliukha3https://orcid.org/0000-0002-3583-7324Higher School of Artificial Intelligence Technologies, Peter the Great St. Petersburg Polytechnic University, Saint Petersburg, RussiaHigher School of Artificial Intelligence Technologies, Peter the Great St. Petersburg Polytechnic University, Saint Petersburg, RussiaHigher School of Artificial Intelligence Technologies, Peter the Great St. Petersburg Polytechnic University, Saint Petersburg, RussiaHigher School of Artificial Intelligence Technologies, Peter the Great St. Petersburg Polytechnic University, Saint Petersburg, RussiaA new method called the Survival Beran-based Neural Importance Model (SurvBeNIM) is proposed. It aims to explain predictions of machine learning survival models, which are in the form of survival or cumulative hazard functions. The main idea behind SurvBeNIM is to extend the Beran estimator by incorporating the importance functions into its kernels and by implementing these importance functions as a set of neural networks which are jointly trained in an end-to-end manner. Two strategies of using and training the whole neural network implementing SurvBeNIM are proposed. The first one explains a single instance, and the neural network is trained for each explained instance. According to the second strategy, the neural network only learns once on all instances from the dataset and on all generated instances. Then the neural network is used to explain any instance in a dataset domain. Various numerical experiments compare the method with different existing explanation methods developed for survival analysis, in particular, for models SurvBex, SurvNAM, and SurvLIME. Results of experiments with synthetic and real data show that the proposed method outperforms existing explanation methods. A code implementing the proposed method is publicly available.https://ieeexplore.ieee.org/document/10858699/Interpretable modelexplainable AILIMEneural additive modelsurvival analysiscensored data |
| spellingShingle | Lev V. Utkin Danila Y. Eremenko Andrei V. Konstantinov Vladimir A. Muliukha SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival Models IEEE Access Interpretable model explainable AI LIME neural additive model survival analysis censored data |
| title | SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival Models |
| title_full | SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival Models |
| title_fullStr | SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival Models |
| title_full_unstemmed | SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival Models |
| title_short | SurvBeNIM: The Beran-Based Neural Importance Model for Explaining Survival Models |
| title_sort | survbenim the beran based neural importance model for explaining survival models |
| topic | Interpretable model explainable AI LIME neural additive model survival analysis censored data |
| url | https://ieeexplore.ieee.org/document/10858699/ |
| work_keys_str_mv | AT levvutkin survbenimtheberanbasedneuralimportancemodelforexplainingsurvivalmodels AT danilayeremenko survbenimtheberanbasedneuralimportancemodelforexplainingsurvivalmodels AT andreivkonstantinov survbenimtheberanbasedneuralimportancemodelforexplainingsurvivalmodels AT vladimiramuliukha survbenimtheberanbasedneuralimportancemodelforexplainingsurvivalmodels |