RULCLIPPER algorithm and CMAPSS health indicators

Prognostics based on computational geometry

Implements an algorithm for "Remaining Useful Life estimation based on impreCise heaLth Indicator modeled by Planar Polygons and similarity-basEd Reasoning" (RULCLIPPER) initially developed for prognostics on CMAPSS datasets. It makes use of elementary polygon operations implemented on Matlab together with recall/precision/f1-measure for similarity estimation. In contrast to its simplicity, it provided very good results on CMAPSS datasets (publication included) compared to many other approaches including neural network or Bayesian (sparse) learning. The package includes a way to estimate health indicators on those datasets (proposed in the publication). Running codes allow to retrieve results of the publication.