Application of Support Vector Machine in Laboratory Information Security Protection

As the information technology develops rapidly, the security of laboratory information systems faces more and more challenges, especially the threats of malicious intrusion and data breach. To effectively identify and prevent these potential security threats, this paper proposes a laboratory information security protection solution based on the support vector machine (SVM) model. By collecting laboratory network traffic, system logs and other data, this paper first preprocesses the data, including noise removal, feature extraction and standardization, and annotates the data to construct a data set suitable for the SVM model. Subsequently, One-Class SVM (OCSVM) is used for anomaly detection. The model learns the distribution of normal data through training and detects whether there is abnormal activity. To further improve the detection precision, this paper gradually optimizes the decision boundary of the model and finally improves the accuracy of the model by adjusting hyperparameters and iterative training. The experimental results indicate that with the increase in the number of iterations, the SVM model shows significant improvement in evaluation indicators such as accuracy, recall, and Fl-score, approaching 100% at the 100th iteration, and the AUC (area under the curve) value of the model is 0.99.