Research on the Application of Logistic Regression Algorithm in Disease Classification Using Steepest Descent Method for Parameter Optimization

In today's era of big data, extracting valuable information from massive medical datasets is increasingly important to enhance the precision and reliability of disease diagnoses made by doctors<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">[1]</sup>. This paper employs the steepest descent method for parameter optimization in disease classification using the logistic regression algorithm. The steepest descent method iteratively updates parameters to gradually minimize the loss function, thereby improving the model's classification performance and aiding in enhancing diagnostic accuracy. The experiment uses an open-source breast cancer dataset, comparing the performance of the logistic regression algorithm with and without parameter optimization via the steepest descent method. Model performance is evaluated through confusion matrices, ROC curves, and learning curves<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">[2]</sup>. Finally, this paper concludes that compared to traditional methods that only use logistic regression algorithms, optimizing the logistic regression algorithm parameters with the steepest descent method significantly improves model classification capabilities in disease classification.