| 摘要 |
<p>AN IMMUNE BASED METHOD FOR DETECTING ANOMALY IN A MONITORED DATA SYSTEM COMPRISING THE STEPS OF ANALYZING A PLURALITY OF HISTORICAL DATA (1) IN PREDETERMINED RULES AND DEFINING THE PLURALITY OF HISTORICAL DATA (1) AS SELF DATA (2); GENERATING A CANDIDATE ELEMENT (4) AND SUBSEQUENTLY MATCHING THE CANDIDATE ELEMENT (4) WITH THE SELF DATA (2) TO ESTABLISH A SET OF MATURED DETECTORS (3), PRESENTING TEST DATA (5) FROM THE MONITORED DATA SYSTEM TO THE DETECTING SYSTEM AND FURTHER MATCHING THE TEST DATA WITH THE SET OF MATURED DETECTORS (3) TO DETECT FOR ABNORMAL DATA BEHAVIOR AND FINALLY PROVIDING A WARNING SIGNAL TO ALERT USER FOR FURTHER ACTION IF AN ABNORMAL DATA BEHAVIOR IS DETECTED. THE IMMUNE BASED SYSTEM OF DETECTING ANOMALY IS DEVELOPED AS A GENERIC DETECTOR TO DETECT A ABNORMAL BEHAVIOR WITHIN A CHAOTIC DATA, SUCH AS STOCK MARKET TRADING DATA, LOAD POWER DATA, RAINFALL DATA AND OTHERS WHEREBY THE SYSTEM IMPLEMENTS THE SYNERGIZED CONCEPT OF ARTIFICIAL IMMUNE SYSTEM AND ROUGH RULE ANALYSIS. THE MOST ILLUSTRATIVE DRAWING:</p> |
