Research On SVM Alarm Signal Recognition Model And Its Algorithm Based On Alarm Combination Prediction

Early warning refers to the degree level of predicting the future deviation from the expected state according to the history and present situation of the object elements,and the degree of deviation is generally expressed by different alarm signals.The Yellow early-warning mechanism is suitable for complex early warning in the economic and social field.For the Yellow early-warning index strategy,if the extrapolation paradigm is used to measure the extrapolation paradigm,the warning signal prediction depends on the level critical point(alarm signal recognition)and the warning of the future period.The degree of critical point alarm is usually selected directly by the group decision,while the future period alarm is generated by the time series prediction method,which is different from the technical way of the formation of the level critical point and the future period.For the concept of abstract critical point alerting,it is difficult for experts to make accurate judgement,making the prediction of the police signal inaccurate and affecting the reliability of early warning.This paper expounds the architecture of the regression support vector machine(SVM)principle and the multiple cross SVM combination prediction of the alarm degree.Based on the critical point of the early warning attribute selected by the group decision,according to the multiple crossover SVM combination prediction calculation and the data trajectory,the warning signal recognition strategy and the technique approach are proposed,and the alarm signal recognition(critical point alarm)is constructed.Degree)model and its algorithm.Taking the alarm signal recognition of the regional eco economic system as an example,the SVM alarm signal recognition model and its algorithm are applied to verify the validity and applicability of the alarm signal recognition method.The positive application shows that the alarm degree generated by the alarm signal recognition method is less than 1.98% compared with the combined prediction,and the recognition alarm has the exact consistency with the combined prediction,the greater the correlation degree between the grade critical point and the time series sample,the more discrepancy between the level critical point and the timing sample combination.Small,there is a correlation difference between the recognition alarm degree and the combination prediction alarm degree.