Machine Learning-based Mixture Bimodal Intelligent Decision System

In the information age,computer technology,network technology and communication technology are widely used.Cultivating and developing computer-based intelligent tools can play a large role in various fields.As the important tools of information acquisition and information conversion,sensor technology has been combined with various disciplines,infiltrating applications in various fields.Electrochemical sensor technology for the detection of material composition has the advantages of high sensitivity,simple and easy operation,low detection cost,and online analysis.It is an important research direction of analytical methods in recent years.Due to the similar structural properties of many substances,electrochemical sensors are likely to interfere with each other in the actual sample detection.Successfully distinguishing these substances has always been the focus of electrochemical detection technology.Two kinds of purines—Adenine(A)and Hypoxanthine(Hx)with similar potentials—were selected as the research object,and a voltammetric detection sensor device was constructed using carbon nanotubes as the modified electrode,simultaneous detection experiments were performed on these two substances and their results were further studied.Two kinds of substances near the peak potential will overlap in the volt-ampere detection experiment.To better analyze the experimental data,it is necessary to study the overlapping peak separation method.The researchers showed that fractional derivatives can effectively separate overlapping peaks.In this paper,we consider that the response peak of adenine and hypoxanthine is the two peak of overlapping,we try to apply the half-order conduction technique to the bimodal separation processing on the graph to achieve the pretreatment of experimental data.After the experimental data was processed by the semi-rank derivative algorithm,the method of graphic comparison was used to prove that the data after the pre-processing of the semi-rank derivative technology can be used.On the basis of the electrochemical sensor testing experiment,this paper combines the detection technology of electrochemical sensor and computer technology,and the(Support Vector Machines)SVM algorithm is used to construct a discriminant system for the simultaneous test results of A and Hx,which is used for the intelligent analysis and processing of the data obtained from the experiment.The system based on the SVM algorithm in this paper is successful in implementing intelligent identification of different test results.After the data processed by the semi-ranking algorithm is imported into the system,intelligent determination of the results of the simultaneous detection of A and Hx is performed to achieve the purpose of intelligent differentiation and reduction of human intervention.In order to make the process simple and efficient in the process of data processing,this paper adopts the K-means algorithm.The K-means algorithm is a typical distance-based clustering algorithm and uses distance as the evaluation index of similarity.In order to evaluate the stability of the constructed sensor in the test experiment,after several tests in this paper,these data were clustered and analyzed with the K-means algorithm to successfully determine the stability of the data.The results show the effectiveness of separation preprocessing,intelligence difference and experimental method stability evaluation method on the data detected simultaneously by A and Hx,which provides a direction for the intellectualization of electrochemical detection.