Development of machine learning models to facilitate sample analysis : prediction of LC retention times and FT-IR image analysis for microplastic detections : Part I. Predict retention time of dansylated metabolites in LC/MS. Part II. Classify the FT-IR spectrum to facilitate microplastic identification.
- 주제어 (키워드) Machine learning , Convolutional neural network , Artificial Neural Network , LC/MS , Dansylation , Metabolomics , Retention time , FT-IR , Microplastic , Prediction
- 발행기관 서강대학교 일반대학원
- 지도교수 오한빈
- 발행년도 2024
- 학위수여년월 2024. 2
- 학위명 박사
- 학과 및 전공 일반대학원 화학과
- 실제URI http://www.dcollection.net/handler/sogang/000000076764
- UCI I804:11029-000000076764
- 본문언어 영어
- 저작권 서강대학교 논문은 저작권 보호를 받습니다.
초록
Today, machine learning mechanisms are evolving dramatically. These methods are being adopted in almost every field, and they are also being adopted in the field of chemistry. This study is part of the introduction of machine learning algorithms in chemistry to facilitate sample analysis. The first part is LC retention time prediction: artificial neural network regression models can be used to predict the retention time of densylated metabolites. The second part is FT-IR image analysis for microplastic detection. This requires reconstructing the image by classifying the FT-IR spectra comprising the image with a one-dimensional convolutional neural network. These are just a few examples of sample analysis using machine learning, specifically artificial intelligence, but soon, these machine learning mechanisms will be more widely used.
more목차
Part I Overview 2
Part II Predict retention time of dansylated metabolites in LC/MS 12
Part III Development of a machine-learning model and software for microplastic analysis in an FT-IR microscopy image 62
