General structure of L-lymphocyte applied to event detection in financial time series

The paper is focused on the T-lymphocyte construction applied to immune-inspired event detection in financial time series. The goal is to recognize symptoms of abrupt of long-time mean value of many processed series. The task of the T-lymphocyte is to distinguish between "healthy" and "illness" states through examining individual series, with algorithms based on weak and rigorous statistical tests (detailed operation of detection is shown). General structure of the L-lymphocyte algorithm is illustrated. A comparison of the number of detected symptoms is presented.