Robust Correction of Sampling Bias using Cumulative Distribution Functions
Authors: Bijan Mazaheri, Siddharth Jain, Jehoshua Bruck
NeurIPS 2020 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Further, we show experimentally that our method is more robust in its predictions, is not reliant on parameter tuning and shows similar classification performance compared to the current state-of-the-art techniques on synthetic and real datasets. |
| Researcher Affiliation | Academia | Bijan Mazaheri Department of Computing and Mathematical Sciences, California Institute of Technology, Pasadena, CA 91125 EMAIL Siddharth Jain Department of Electrical Engineering, California Institute of Technology, Pasadena, CA, 91125 EMAIL Jehoshua Bruck Department of Electrical Engineering, California Institute of Technology, Pasadena, CA, 91125 EMAIL |
| Pseudocode | Yes | Algorithm 1 Covariate Shift Classification 1 [...] Algorithm 2 Covariate Shift Classification 2 |
| Open Source Code | No | The paper does not provide any links to source code for their proposed method, nor does it state that the code is available in supplementary materials or elsewhere. |
| Open Datasets | Yes | We conduct experiments on real datasets [18, 20, 9, 13, 21] and show comparable performance to other widely known covariate shift methods [11, 5, 14, 15, 6]. [...] twonorm and ringnorm datasets [9]. The cancer, diabetes and banknote datasets. |
| Dataset Splits | Yes | N = 200 training points and M = 1000 testing points were used. [...] A training set of size 100 is chosen uniformly at random from the remaining samples in the dataset. [...] The mean performance error over 100 trials for twonorm and ringnorm datasets [9] is provided in Table 1. |
| Hardware Specification | No | The paper does not specify any hardware details (e.g., GPU models, CPU types, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions "matlab code on the authors website" for other methods but does not provide specific software dependencies with version numbers for their own implementation. |
| Experiment Setup | Yes | In Experiments 1 and 2 below, a SVM is used with a square rooted Gaussian kernel with kernel width = 1 and regularization coefficient γ = 0.1. |