Application of Data Mining Technique to Identify the Relationships of Potential Factors of Bogie Instability Alarm in High Speed Train

[[abstract]]本研究利用資料探勘之決策樹演算法，針對間歇性之輪軌介面問題，尋找轉向架異常晃動告警之關鍵構成因子及其構成順序。 依決策樹觀念而言，其所歸納之規則有其構成順序，當一個屬性消失時，則規則無法成立；依此原則下進一步探討相關規則並提出預防性維修建議。(紙本延後至108715止公開)[[abstract]]Decision Tree algorithmic is proposed to study the potential factors of wheel/rail interface concerns dig out of relationships of potential factors and rules of BIDS alarm is expected In light of the concept of Decision Tree the rule has its order While attributes are taken out from the rule the rule can’t be established In that basis recommendation of Preventive Maintenance can be advised accordingl