Deep self-supervised machine learning algorithms with a novel feature elimination and selection approaches for blood...

Area of Science:

• Biomedical data analysis
• Machine learning in healthcare

Background:

• Blood tests are crucial for health screening, diagnosis, and surveillance.
• Automated analysis of blood test data using deep self-supervised machine learning is under-investigated.

Purpose of the Study:

• To develop and evaluate deep machine learning algorithms for automated health risk classification from blood test data.
• To improve the efficiency and accuracy of interpreting complex blood test results.

Main Methods:

• Proposed novel deep machine learning algorithms incorporating multi-dimensional adaptive feature elimination.
• Implemented self-feature weighting and advanced feature selection techniques.
• Modified four distinct machine learning algorithms for health risk classification.

Main Results:

• The developed algorithms successfully identified and removed irrelevant features from blood test data.
• Self-importance weighting and feature selection enhanced the informativeness of the processed data.
• Accurate classification of health risks was achieved across a spectrum of severity.

Conclusions:

• The proposed deep machine learning approach enables automated and accurate health risk classification from blood test data.
• These algorithms demonstrate efficacy in feature selection and weighting for improved diagnostic insights.
• This work paves the way for more sophisticated automated interpretation of clinical laboratory results.