Transfer learningIn this work, we study the transfer learning problem under high-dimensional generalized linear models (GLMs), which aim to improve the fit on target data by borrowing information from useful source data. Given which sources to transfer, we propose a transfer learning algorithm on ...
Li, S., Cai, T. T., & Li, H. (2022). Transfer learning for high-dimensional linear regression: Prediction, estimation and minimax optimality. Journal of the Royal Statistical Society Series B: Statistical Methodology, 84, 149–173. Zhao J, Zheng S, Leng C. Residual Importance Weighted T...
This paper studies transfer learning of a high-dimensional generalized linear model with the target model as well as source data from different but possibly related models. Both known and unknown transferable domain settings are considered. On the one hand, an improved two-step transfer learning alg...
Transfer learning under high-dimensional generalized linear models. Tian, Y., & Feng, Y. (2022). Journal of the American Statistical Association, (pp. 1–14). Transfer Learning for Functional Linear Regression with Structural Interpretability. Haotian Lin and Matthew Reimherr Residual Importance Weig...
Transfer learning is amachine learningapproach that involves utilizing knowledge acquired from one task to improve performance on a different but related task. For example, if we train a model to recognize backpacks in pictures, we can use it to identify objects like sunglasses, a cap or a tabl...
We find that the mapping from gene expression patterns to cell types can be learnt using a simple generalized linear model (here, MLR) that is easily interpretable. This learning is only marginally improved by more complex artificial neural networks, which suffer from poor interpretability. Our res...
(Fig.3). We have observed through experimental validation that the ANN outperforms linear regression model when used on batteries of unseen devices. Additionally, transfer learning is possible using ANNs, where, the model generalizes on new data while retaining its earlier learning. The ANN has ...
We establish non-asymptotic recovery guarantees for the exttt{$\ell_1$-TCL} with generalized linear model (GLM) under the sparsity assumption in the high-dimensional setting, and demonstrate the empirical benefits of exttt{$\ell_1$-TCL} through extensive numerical simulation for GLM and recent...
The first work on causal transfer learning 日本理论组大佬Sugiyama的工作,causal transfer learning 20191008 CVPR-19 Characterizing and Avoiding Negative Transfer Characterizing and avoid negative transfer 形式化并提出如何避免负迁移 20190301 ALT-19 A Generalized Neyman-Pearson Criterion for Optimal Dom...
(fig. 1 a). the degas framework in its simplest form can be broken into three tasks during model training: (1) correctly labeling cells with a cellular subtype using multitask learning; (2) correctly assigning clinical labels to patients using multitask learning; and (3) generating a latent...