Negative and positive classification model establishment method and device, equipment and computer storage medium
A technology for classifying models and establishing methods, which is applied in computer parts, calculation, image analysis, etc., can solve problems such as irregular dPCR sample brushing, false negative amplification efficiency, non-specific probe hybridization, etc., to ensure dPCR quantification, The effect of saving labor costs and high classification accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0079] An embodiment of the present invention provides a method for establishing a negative-positive classification model based on a real-time digital PCR system, which is applied to a dPCR system whose heat dissipation efficiency is lower than a preset value, see figure 2 shown, including:
[0080] Step S12, for a single dPCR amplification reaction, respectively select the first number of negative samples and positive samples;
[0081] Step S14, respectively selecting the second number of negative samples and positive samples as training samples and inputting them into the preset SVM training model out of sequence, to obtain the first hyperplane model that meets the preset requirements;
[0082] Step S16, taking the third number of negative samples and positive samples respectively as test samples, and sequentially inputting the first hyperplane model, and determining the first hyperplane model when the correct rate of the category of the output test samples reaches a set th...
Embodiment 2
[0133] An embodiment of the present invention provides a method for establishing a negative-positive classification model. Based on a real-time digital PCR system, it is applied to a dPCR system whose heat dissipation efficiency is lower than a preset value, and can realize negative, positive, first false positive and second positive results for test samples. Four categories of false positives, see Figure 5 shown, including:
[0134] Step S22, for a single dPCR amplification reaction, respectively select the first number of negative samples and positive samples, and the fourth number of first false positive samples and second false positive samples;
[0135] Step S24, respectively select the second number of negative samples and positive samples, and the fifth number of first false positive samples and second false positive samples as training samples and input them into the preset SVM training model in random order, and obtain the number that meets the preset requirements. ...
Embodiment 3
[0152] Corresponding to the method for establishing a negative-positive classification model for binary classification, an embodiment of the present invention provides a device for establishing a negative-positive classification model based on a real-time digital PCR system, which is applied to a dPCR system whose heat dissipation efficiency is lower than a preset value, see Figure 8 Shown include:
[0153] The first sample selection module 81 is used to select a first number of negative samples and positive samples respectively for a single dPCR amplification reaction;
[0154] The first hyperplane model obtaining module 82 selects the negative samples and positive samples of the second quantity respectively as training samples and inputs the preset SVM training model out of order to obtain the first hyperplane model that meets the preset requirements;
[0155] The first verification module 83 is used to respectively use the third number of negative samples and positive samp...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com