Outlier Detection Algorithms Over Fuzzy Data with Weighted Least Squares
نویسندگان
چکیده
In the classical leave-one-out procedure for outlier detection in regression analysis, we exclude an observation and then construct a model on remaining data. If difference between predicted observed value is high declare this outlier. As rule, those procedures utilize single comparison testing. The problem becomes much harder when observations can be associated with given degree of membership to underlying population, should generalized operate over fuzzy We present new approach that operates data using two inter-related algorithms. Due way outliers enter sample, they may various order magnitude. To account this, divided into cycles. Furthermore, each cycle consists phases. Phase 1, apply non-outlier dataset. 2, all previously declared are subjected Benjamini–Hochberg step-up multiple testing controlling false-discovery rate, non-confirmed return Finally, resulting set non-outliers. way, ensure reliable high-quality obtained 1 because comparatively easily purges dubious due same time, confirmation status relation newly applied hence only true remain outside sample. phases good trade-off desire (i.e., informative points) use as points possible (thus leaving sample). number cycles user defined, but finalize analysis case no detected. offer one illustrative example other practical studies (from real-life thrombosis studies) demonstrate application strengths our concluding section, discuss several limitations also directions future research.
منابع مشابه
Outlier detection algorithms for least squares time series regression 1
We review recent asymptotic results on some robust methods for multiple regression. The regressors include stationary and non-stationary time series as well as polynomial terms. The methods include the Huber-skip M-estimator, 1-step Huber-skip M-estimators, in particular the Impulse Indicator Saturation, iterated 1-step Huber-skip M-estimators and the Forward Search. These methods classify obse...
متن کاملPEDOMODELS FITTING WITH FUZZY LEAST SQUARES REGRESSION
Pedomodels have become a popular topic in soil science and environmentalresearch. They are predictive functions of certain soil properties based on other easily orcheaply measured properties. The common method for fitting pedomodels is to use classicalregression analysis, based on the assumptions of data crispness and deterministic relationsamong variables. In modeling natural systems such as s...
متن کاملWeighted Majorization Algorithms for Weighted Least Squares Decomposition Models
For many least-squares decomposition models efficient algorithms are well known. A more difficult problem arises in decomposition models where each residual is weighted by a nonnegative value. A special case is principal components analysis with missing data. Kiers (1997) discusses an algorithm for minimizing weighted decomposition models by iterative majorization. In this paper, we propose a m...
متن کاملFuzzy least-squares algorithms for interactive fuzzy linear regression models
Fuzzy regression analysis can be thought of as a fuzzy variation of classical regression analysis. It has been widely studied and applied in diverse areas. In general, the analysis of fuzzy regression models can be roughly divided into two categories. The 0rst is based on Tanaka’s linear-programming approach. The second category is based on the fuzzy least-squares approach. In this paper, new t...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: International Journal of Fuzzy Systems
سال: 2021
ISSN: ['2199-3211', '1562-2479']
DOI: https://doi.org/10.1007/s40815-020-01049-8