Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation
Near infrared spectroscopic (NIRS) data from different harvested seasons may consist of different feature spaces even though the samples have the same label values. This is because the spectral response could be affected by the changes in environmental parameters, internal quality, and the reprod...
| Main Authors: | , |
|---|---|
| Format: | Book Section |
| Language: | English |
| Published: |
Springer
2022
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/7593/ http://eprints.uthm.edu.my/7593/1/C3879_ae4cee62c76825b1343f870777ce3a0e.pdf |
| _version_ | 1848889148959621120 |
|---|---|
| author | Suarin, Nur Aisyah Syafinaz Kim, Seng Chia |
| author_facet | Suarin, Nur Aisyah Syafinaz Kim, Seng Chia |
| author_sort | Suarin, Nur Aisyah Syafinaz |
| building | UTHM Institutional Repository |
| collection | Online Access |
| description | Near infrared spectroscopic (NIRS) data from different harvested
seasons may consist of different feature spaces even though the samples have the
same label values. This is because the spectral response could be affected by the
changes in environmental parameters, internal quality, and the reproducibility of
NIRS instruments. Thus, this study aims to investigate the ability of Joint
Distribution Adaptation (JDA) transfer learning algorithm in addressing the
assumption of traditional machine learning i.e. both training and testing data
must come from the same feature spaces and data distribution. First, NIRS data
acquired from two different harvested seasons were used as the source domain
and the target domain, respectively. Next, JDA was implemented to produce an
adaptation matrix using the source domain and transfer datasets. This adaptation
matrix would be used to transform the source and target domain datasets. After
that, a calibration model was developed by means of Partial Least Squares
(PLS) using the transformed training dataset; and validated using the trans�formed independent testing dataset. The proposed JDA-PLS was compared to
the PLS without transfer learning as the baseline learning. Findings show that
the proposed JDA-PLS with 10 LVs achieved the lowest RMSEP of 1.134% and
the highest RP
2 of 0.826. |
| first_indexed | 2025-11-15T20:21:35Z |
| format | Book Section |
| id | uthm-7593 |
| institution | Universiti Tun Hussein Onn Malaysia |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-15T20:21:35Z |
| publishDate | 2022 |
| publisher | Springer |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | uthm-75932022-08-29T07:35:33Z http://eprints.uthm.edu.my/7593/ Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation Suarin, Nur Aisyah Syafinaz Kim, Seng Chia T Technology (General) Near infrared spectroscopic (NIRS) data from different harvested seasons may consist of different feature spaces even though the samples have the same label values. This is because the spectral response could be affected by the changes in environmental parameters, internal quality, and the reproducibility of NIRS instruments. Thus, this study aims to investigate the ability of Joint Distribution Adaptation (JDA) transfer learning algorithm in addressing the assumption of traditional machine learning i.e. both training and testing data must come from the same feature spaces and data distribution. First, NIRS data acquired from two different harvested seasons were used as the source domain and the target domain, respectively. Next, JDA was implemented to produce an adaptation matrix using the source domain and transfer datasets. This adaptation matrix would be used to transform the source and target domain datasets. After that, a calibration model was developed by means of Partial Least Squares (PLS) using the transformed training dataset; and validated using the trans�formed independent testing dataset. The proposed JDA-PLS was compared to the PLS without transfer learning as the baseline learning. Findings show that the proposed JDA-PLS with 10 LVs achieved the lowest RMSEP of 1.134% and the highest RP 2 of 0.826. Springer 2022 Book Section PeerReviewed text en http://eprints.uthm.edu.my/7593/1/C3879_ae4cee62c76825b1343f870777ce3a0e.pdf Suarin, Nur Aisyah Syafinaz and Kim, Seng Chia (2022) Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation. In: Control, Instrumentation and Mechatronics: Theory and Practice. Springer, pp. 707-716. |
| spellingShingle | T Technology (General) Suarin, Nur Aisyah Syafinaz Kim, Seng Chia Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation |
| title | Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation |
| title_full | Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation |
| title_fullStr | Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation |
| title_full_unstemmed | Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation |
| title_short | Transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation |
| title_sort | transferring near infrared spectroscopic calibration model across different harvested seasons using joint distribution adaptation |
| topic | T Technology (General) |
| url | http://eprints.uthm.edu.my/7593/ http://eprints.uthm.edu.my/7593/1/C3879_ae4cee62c76825b1343f870777ce3a0e.pdf |