SVR在传感器预测中的研究

The SVR in sensor prediction research

环境因素产生的交叉敏感使得传感器在精度上往往难以满足,文章介绍了将支持向量机回归引入到传感器,较好的泛化能力使得输出逼近于期望值。研究结果表明,支持向量机的引入在降低交叉敏感度有明显的效果,从而证明了该方法的可行性。

Cross sensitivity produced by environmental factors will reduce the precision of the sensor.This paper introduces SVM to sensor, because of good generalization ability make the sensor output close to hope value. The result shows the obvious effect of this method and proofed the the feasibility of this method.