1: \begin{abstract}

2: Feature selection places an important role in improving the performance of outlier detection, especially for noisy data. 

3: Existing methods usually perform feature selection and outlier scoring separately, which would select feature subsets that may not optimally serve for outlier detection, leading to unsatisfying performance.

4: In this paper, we propose an outlier detection ensemble framework with embedded feature selection (ODEFS), to address this issue. 

5: Specifically, for each random sub-sampling based learning component, ODEFS unifies feature selection and outlier detection into a pairwise ranking formulation to learn feature subsets that are tailored for the outlier detection method. 

6: Moreover, we adopt the thresholded self-paced learning to simultaneously optimize feature selection and example selection, which is helpful to improve the reliability of the training set. 

7: After that, we design an alternate algorithm with proved convergence to solve the resultant optimization problem. 

8: In addition, we analyze the generalization error bound of the proposed framework, which provides theoretical guarantee on the method and insightful practical guidance. 

9: Comprehensive experimental results on 12 real-world datasets from diverse domains validate the superiority of the proposed ODEFS.

10: \end{abstract}

11: