?url_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=Contextual+normalization+applied+to+aircraft+gas+turbine+engine+diagnosis&rft.creator=Turney%2C+Peter&rft.creator=Halasz%2C+Michael&rft.subject=Artificial+Intelligence&rft.subject=Machine+Learning&rft.subject=Statistical+Models&rft.description=Diagnosing+faults+in+aircraft+gas+turbine+engines+is+a+complex+problem.+It+involves+several+tasks%2C%0Aincluding+rapid+and+accurate+interpretation+of+patterns+in+engine+sensor+data.+We+have+investigated%0Acontextual+normalization+for+the+development+of+a+software+tool+to+help+engine+repair+technicians%0Awith+interpretation+of+sensor+data.+Contextual+normalization+is+a+new+strategy+for+employing%0Amachine+learning.+It+handles+variation+in+data+that+is+due+to+contextual+factors%2C+rather+than+the%0Ahealth+of+the+engine.+It+does+this+by+normalizing+the+data+in+a+context-sensitive+manner.+This%0Alearning+strategy+was+developed+and+tested+using+242+observations+of+an+aircraft+gas+turbine%0Aengine+in+a+test+cell%2C+where+each+observation+consists+of+roughly+12%2C000+numbers%2C+gathered+over+a%0A12+second+interval.+There+were+eight+classes+of+observations%3A+seven+deliberately+implanted+classes%0Aof+faults+and+a+healthy+class.+We+compared+two+approaches+to+implementing+our+learning+strategy%3A%0Alinear+regression+and+instance-based+learning.+We+have+three+main+results.+(1)+For+the+given%0Aproblem%2C+instance-based+learning+works+better+than+linear+regression.+(2)+For+this+problem%2C%0Acontextual+normalization+works+better+than+other+common+forms+of+normalization.+(3)+The%0Aalgorithms+described+here+can+be+the+basis+for+a+useful+software+tool+for+assisting+technicians+with%0Athe+interpretation+of+sensor+data.&rft.date=1993&rft.type=Journal+(Paginated)&rft.type=PeerReviewed&rft.format=application%2Fpdf&rft.identifier=http%3A%2F%2Fcogprints.org%2F1864%2F3%2FTurbine.pdf&rft.identifier=++Turney%2C+Peter+and+Halasz%2C+Michael++(1993)+Contextual+normalization+applied+to+aircraft+gas+turbine+engine+diagnosis.++%5BJournal+(Paginated)%5D+++++&rft.relation=http%3A%2F%2Fcogprints.org%2F1864%2F