Research on Intelligent Algorithm of Cancer Diagnosis Evaluation Based on Machine Learning

Scientific and efficient intelligent classification model is an important basis for cancer evaluation and diagnosis. Traditional cancer diagnosis and evaluation methods are greatly affected by human subjective factors. Support vector machine can classify linearly indivisible data sets very well, but it is easy to misclassify sample points that are close to the classification hyperplane. The nearest neighbor algorithm can also classify the samples well, and it is not sensitive to outliers in the classification process, but the calculation is large. According to the advantages and disadvantages of support vector machine model and nearest neighbor model, the sample points that are close to the classification hyperplane in support vector machine are classified by the nearest neighbor model, that is, the SVM-KNN combination model is proposed, so as to train an intelligent classification diagnosis model with high classification accuracy and is effective and feasible. A breast cancer data set was used for empirical analysis, and the combined model improved the classification accuracy of the data set.

I. Introduction

Cancer is one of the most serious diseases threatening human life in the 21st century. It has the characteristics of rapid onset, high fatality rate and difficult treatment, so people think that cancer is incurable. With the development of medical technology, cancer has become not incurable as people think, and the World Health Organization has clearly pointed out that the important ways to prevent cancer are: Early detection, early diagnosis, early treatment, therefore, it is very important for early cancer patients to be diagnosed as early as possible, therefore, accurate cancer screening, diagnosis, evaluation methods and scientific and efficient treatment methods are the key to the treatment of cancer this disease. There is an increasing amount of information and data related to the cause of cancer. Although these massive data help the accuracy of cancer diagnosis, it also brings greater challenges to the full utilization and analysis of these data.