BLUF: A new method for detecting outliers in linear-circular non-parametric regression based on circular distances from the median value for Wrapped-Cauchy distributed data is proposed, with performance supported by a real dataset and simulation study with varying contamination and sample size. Non-parametric Nadaraya-Watson and local linear regression methods were employed to obtain regression fits, with LL providing better fit when the response variable contained outliers.
Subscribe
Login
Please login to comment
0 Comments
Most Voted