1: \begin{abstract}

2: In additive models with many nonparametric components, 

3: a number of regularized estimators have been proposed and proven to attain various  

4: error bounds under different combinations of sparsity and fixed smoothness conditions.  

5: Some of these error bounds match minimax rates in the corresponding settings. 

6: Some of the rate minimax methods are non-convex and %thus 

7: computationally costly. 

8: From these perspectives, the existing solutions to the high-dimensional 

9: additive nonparametric regression problem are fragmented. 

10: In this paper, we propose a multi-resolution group Lasso (MR-GL) method in a unified approach 

11: to simultaneously achieve or improve existing error bounds and provide new ones 

12: without the knowledge of the level of sparsity or the degree of smoothness of the unknown functions. 

13: Such adaptive convergence rates are established when a prediction factor can be treated as a constant. 

14: Furthermore, we prove that the prediction factor, which can be bounded in terms of a restricted eigenvalue or a compatibility coefficient, can be indeed treated as a constant 

15: for random designs under a nearly optimal sample size condition. 

16: %\chzR{This unified approach is taken in a more general additive model with a large 

17: %number of components of mixed types, including univariate linear effects, 

18: %group linear effects and nonparametric effects of one or several variables. 

19: %A prediction error bound is provided in the general additive model 

20: %to achieve desired convergence rates when a prediction factor can be treated as a constant. 

21: %The prediction factor, which can be bounded in terms of a restricted eigenvalue or a compatibility coefficient, is indeed bounded under proper conditions for random designs.} 

22: \end{abstract}

23: