Generalized Additive Model Using Deep Neural Networks with Self-attention Module
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The generalized additive model (GAM) is a flexible statistical tool adept at unraveling complex relationships in datasets. Deep neural networks (DNNs) stand out as robust and versatile models that perform exceptionally well across various computational challenges. In this study, we introduce DeepGAM, an innovative model that leverages the feature-learning capabilities of DNNs while maintaining the interpretability inherent to GAM. This hybrid model is designed to learn a linear combination of neural networks, with each network focusing on a single predictor variable and incorporating a self-attention mechanism to highlight critical features. By training the networks in unison, DeepGAM proficiently maps the nuanced connections between predictors and outcomes. By conducting a comprehensive numerical analysis that encompasses regression and classification tasks, the efficiency and versatility of DeepGAM in addressing diverse analytical problems are validated.
Abstract
The generalized additive model (GAM) is a flexible statistical tool adept at unraveling complex relationships in datasets. Deep neural networks (DNNs) stand out as robust and versatile models that perform exceptionally well across various computational challenges. In this study, we introduce DeepGAM, an innovative model that leverages the feature-learning capabilities of DNNs while maintaining the interpretability inherent to GAM. This hybrid model is designed to learn a linear combination of neural networks, with each network focusing on a single predictor variable and incorporating a self-attention mechanism to highlight critical features. By training the networks in unison, DeepGAM proficiently maps the nuanced connections between predictors and outcomes. By conducting a comprehensive numerical analysis that encompasses regression and classification tasks, the efficiency and versatility of DeepGAM in addressing diverse analytical problems are validated.
참고문헌 (12)
- Hastie T, "Generalized additive models", pp.1990, Chapman &Hall
- Nelder J, "Generalized linear models", J R Stat Soc Ser A, 135, pp.370~384 (1972)
- Wasserman L, "All of nonparametric statistics" (2006), Springer Publisher
- Tsybakov AB, "Introduction to nonparametric estimation" (2008), Springer Publisher
- Cortes C, "Support vector networks", Mach Learn, 24, pp.273~297 (1995)
- Cho Y, "Binding affinity prediction using selfsupervised learning and QP-Ensemble", QBS, 42, pp.57~64 (2023)
- Shim J, "Varying coefficient regression modeling using deep neural network", QBS, 41, pp.71~76 (2022)
- Agarwal R, "Neural additive models: interpretable machine learning with neural nets", Advances in Neural Information Processing Systems, 33(NeurIPS), pp.4699~4711 (2021)
- Vaswani A, "Attention is all you need", Advances in Neural Information Processing Systems (NIPS), pp.5998~6008 (2017)
- Harrison D, "Hedonic prices and the demand for clean air", J Environ Econ Manag, 5, pp.81~102 (1979)
- Chang V, "Pima Indians diabetes mellitus classification based on machine learning (ML) algorithms", Neural Comput Appl, 35, pp.16157~16173 (2022)