"تشخیص دیابت با یادگیری ماشین: رویکردی خودتوضیحی و با دسترسی آسان"
کد: G-1656
نویسندگان: Alireza Naraki * ℗
زمان بندی: زمان بندی نشده!
برچسب: سیستم های تصمیم یار بالینی
دانلود: دانلود پوستر
خلاصه مقاله:
خلاصه مقاله
Background and Objectives: Diagnosis of diabetes is a critical area of medical research, as uncontrolled diabetes can lead to severe complications such as heart disease, damage to blood vessels, kidneys, eyes, and nerves. The aim of this study is to improve the accuracy and efficiency of diabetes diagnosis using advanced statistical techniques and machine learning, as well as to create an easily accessible and self-explanatory user interface. This user interface not only helps doctors and patients to receive diagnosis results, but also allows for the collection of new data from users. Method: In this study, a combination of machine learning models, including XGBoost, LightGBM, Gradient Boosting, and CatBoost, is used to diagnose diabetes based on the Pima Indians dataset. To enhance performance, hybrid techniques and statistical methods are used to increase the number of training samples. Results: The proposed method achieves an accuracy of 87%, which represents an improvement of approximately 1–4% compared to standard methods and recent studies. In addition, these techniques offer advantages such as faster learning processes and reduced need for training data. Conclusion: To improve usability, an **explainable and easily accessible user interface** has been developed, which is accessible via a website. This interface enables clinicians and users to upload data, receive diagnostic results, and understand the model's decisions through clear explanations. Integrating machine learning with an interactive user interface can significantly aid medical decision-making.
کلمات کلیدی
Diabetes Diagnosis, Advanced Statistical Techniques, CatBoost
