Gynostemma pentaphyllum seed oil activity prediction system and method based on multi-modal data fusion

By using multimodal data fusion technology, a fused feature vector of Gynostemma pentaphyllum seed oil was constructed, which solved the problem of the ineffective fusion of spectral and chemical data in traditional methods, and improved the accuracy of Gynostemma pentaphyllum seed oil activity prediction and the prediction effect of the model.

CN121703033BActive Publication Date: 2026-06-19SICHUAN HEALTH REHABILITATION VOCATIONAL COLLEGE
2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SICHUAN HEALTH REHABILITATION VOCATIONAL COLLEGE
Filing Date
2026-02-09
Publication Date
2026-06-19

AI Technical Summary

⚠Technical Problem

Traditional activity prediction methods fail to effectively integrate near-infrared spectral data, chemical composition data, and in vitro activity data, resulting in limited analytical capabilities for the complex system of Gynostemma pentaphyllum seed oil and an inability to accurately predict its activity.

⚗Method used

By using multimodal data fusion technology, near-infrared spectral data, chemical composition data, and in vitro activity data of Gynostemma pentaphyllum seed oil were obtained. Normal transformation, missing value imputation, dimensional normalization, and sample alignment were performed to construct a primary fusion feature set. Feature weights were then assigned and fused using a partial least squares regression model to obtain the fusion feature vector of Gynostemma pentaphyllum seed oil. Finally, regression prediction was performed.

🎯Benefits of technology

The multimodal fusion of spectral, bioactivity, and chemical characteristics was achieved, which improved the accuracy and precision of Gynostemma pentaphyllum seed oil bioactivity prediction and enhanced the model's fitting efficiency and prediction performance.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This application provides a system and method for predicting the activity of Gynostemma pentaphyllum seed oil based on multimodal data fusion, relating to the field of multimodal data fusion technology. The method involves performing a normal transformation on the near-infrared spectral data of Gynostemma pentaphyllum seed oil to obtain its spectral principal component features; imputing missing values ​​and normalizing the dimensions of its chemical components to obtain its chemical distribution features, thus determining a primary fusion feature set; fusing all feature vectors in the primary fusion feature set based on the activity scaling values ​​in the activity reference vector of Gynostemma pentaphyllum seed oil, resulting in a fusion feature vector; and performing regression prediction on the activity of Gynostemma pentaphyllum seed oil based on this fusion feature vector to obtain a predicted activity value, which is then output as the prediction result. Based on this scheme, multimodal fusion of spectral features, activity features, and chemical features can be achieved.
Need to check novelty before this filing date? Find Prior Art