Early diabetes risk prediction method based on deep PCA transformation

Abstract

The invention belongs to the technical field of data processing, and particularly relates to an early diabetes risk prediction method based on deep PCA transformation, which comprises the following steps: inputting an early diabetes data set; preprocessing the data, calculating a Pearson correlation coefficient, and filtering redundant features to obtain input data; extracting a feature set of input data through depth PCA to serve as input of a training logistic regression classifier; training a logistic regression classifier based on the feature set for judging a to-be-evaluated case sample; inputting new case sample information, and outputting a result of judging whether the sample suffers from diabetes or not and a corresponding confidence coefficient. Effective extraction of case sample binarization information is realized through a deep PCA-based feature transformation method, and meanwhile, a logistic regression classifier is established to realize judgment of a diseased sample and output a confidence quantitative index of a result, so that early auxiliary diagnosis of an existing diabetes case is conveniently and effectively realized, and an illness state is discovered in time.

Description

technical field [0001] The invention belongs to the technical field of data processing, and in particular relates to an early diabetes risk prediction method based on deep PCA transformation. Background technique [0002] According to the report of the World Health Organization in 2018, diabetes is one of the fastest growing chronic life-threatening diseases, which has affected 422 million people worldwide. Since diabetes has a relatively long asymptomatic period, about 50% of diabetic patients are due to Prolonged asymptomatic periods without timely diagnosis, however early detection of diabetes is important for patient management and is only possible through proper assessment of common and less common signs and symptoms Different stages were found. [0003] Reasons for problems or defects: Currently, the diagnosis of diabetes basically relies on clinicopathological analysis, which cannot effectively realize the early diagnosis of diabetes. Contents of the invention [...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above-mentioned technical problem that the diagnosis of diabetes basically relies on clinicopathological analysis and cannot effectively realize the early diagnosis of diabetes, the present invention provides an early diabetes risk prediction method based on deep PCA transformation that is easy to use, high in efficiency and high in accuracy

Images