Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

CNN and model fusion based blood glucose prediction method for diabetes

A technology of model fusion and prediction methods, applied in the field of machine learning, can solve problems such as diabetes that cannot be cured, and achieve good prediction results

Inactive Publication Date: 2019-03-08
DALIAN UNIV
View PDF14 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, diabetes cannot be cured, and the incidence can only be reduced through scientific and effective prevention

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • CNN and model fusion based blood glucose prediction method for diabetes
  • CNN and model fusion based blood glucose prediction method for diabetes
  • CNN and model fusion based blood glucose prediction method for diabetes

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0014] In order to describe the present invention more specifically, the technical solutions of the present invention will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0015] The implementation of the present invention uses jupyter notebook as the development platform and python as the development language. The data used comes from the Tianchi data platform with a total of 6642 patient physical examination data, and each data has 40 physical examination indicators. The evaluation of the results uses the mean square error (MSE) as the evaluation index of the prediction effect.

[0016]

[0017] The specific steps of the diabetes blood sugar prediction method based on CNN and model fusion of the present invention are as follows:

[0018] S1 preprocessing of data:

[0019] (1) Handling of blank values:

[0020] 1) For indicators with a large number of missing data, such as hepatitis B e antibody, hepatitis...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a CNN and model fusion based blood glucose prediction method for diabetes. The method includes the following steps: performing pre-processing, including the processing of vacancy values, the binarization of quantitative features and data conversion, on data; performing feature extraction on pre-processed data by utilizing a CNN; and performing model fusion on xgboost, catboost and linearRegression by utilizing a Stacking strategy. Catboost and xgboost models with high classification precision are adopted at a first layer, linearRegression with good robustness is adopted in a second layer model, so that the unification of prediction precision and model robustness can be realized, and stronger generalization abilities can be achieved. Thus, the blood glucose prediction problem of diabetes can be effectively solved, and compared with traditional blood glucose prediction methods, the method has significant improvement.

Description

technical field [0001] The invention relates to the field of machine learning, in particular to feature extraction of structured data by using a convolutional neural network (CNN) and prediction of blood sugar by using model fusion technology. Background technique [0002] In recent years, the prevalence of diabetes has been increasing worldwide, and it has become the third major non-communicable disease that threatens human health after tumors and cardiovascular and cerebrovascular diseases. At present, diabetes cannot be cured, and the incidence can only be reduced through scientific and effective prevention. Since blood glucose concentration often has some internal connection with other physical examination indicators, it is helpful to use machine learning technology to establish a blood glucose concentration prediction model to accurately assess people's disease risk based on other physical examination indicators, and then to intervene in high-risk individuals. for effe...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G16H50/50G16H50/20G06N3/04
CPCG16H50/20G16H50/50G06N3/045
Inventor 车超赵撼宇
Owner DALIAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products