| Summary: | The detection of outlier curves/images is crucial in many areas, such as environmental, meteorological, medical, or economic contexts. In the functional framework, outlying observations are not only those that contain atypically high or low values, but also curves that present a different shape or pattern from the rest of the curves in the sample. In this short paper, we mention some recent methods for outlier detection in functional data and apply a recently proposed measure, the directional outlyingness, and the functional outlier map to detect words with outlying distance distribution in the human genome.
|
|---|