基于ν-SVR的纸基纳米金检测Cr6+的浓度识别

Nano-goldbased membranes have been proven to be rapid, precise and selective colorimetric sensors for assay of Cr6+, the intensity of variation of the membranes' color can express the concentration of Cr6+. To address the deficiencies caused by manual analysis, such as inefficiency, instability, low-repeatability, etc. A method based on support vector regression (SVR) is proposed in the paper. Specifically, choose the appropriate training data and test data, use k-fold cross-validation to choose optimal parmeters to complete model training, then use the trained model to identify the concentration. Extensive experiments demonstrate that the identification accuracy of the proposed method is higher than polynomial nonlinear regression and BP neural network, the method can achieve precise concentrationidentification.