Lung fibrosis prediction device based on ensemble learning

Abstract

The invention relates to a lung fibrosis prediction device based on integrated learning. The lung fibrosis prediction device comprises a training sample construction module used for constructing a training sample; the sub-sample construction module is used for randomly extracting n sub-samples from the training sample; the sub-model generation module is used for training the n sub-samples by using a classifier to generate n sub-models; the lung fibrosis prediction model construction module is used for constructing a lung fibrosis prediction model based on the n sub-models; and the lung fibrosis prediction module is used for predicting the lung fibrosis condition of the patient according to the lung fibrosis prediction model. The lung fibrosis condition can be effectively predicted.

Description

technical field [0001] The invention relates to the field of auxiliary medical technology, in particular to a device for predicting pulmonary fibrosis based on integrated learning. Background technique [0002] Pulmonary fibrosis is the end-stage change of a large class of lung diseases characterized by fibroblast proliferation and accumulation of a large amount of extracellular matrix, accompanied by inflammatory damage and tissue structure destruction, that is, normal alveolar tissue is damaged and undergoes abnormal repair to lead to structural changes. Abnormal (scarring). The cause of most patients with pulmonary fibrosis is unknown (idiopathic). This group of diseases is called idiopathic interstitial pneumonia (IIP), which is a large category of interstitial lung diseases. Among idiopathic interstitial pneumonia (IIP), the most common disease type with pulmonary fibrosis as the main manifestation is idiopathic pulmonary fibrosis (IPF), which is a serious disease that...

AI Technical Summary

Problems solved by technology

[0004] However, the existing ensemble learning models do not optimize categorical variables. The way to deal with categorical variables is to add new variables and assign values ​​of 0 or 1, which makes the number of independent variables too large, which easily leads to overfitting problems. thereby reducing the prediction accuracy of the model

Images