Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Analytical-Chemistry-Informed Transformer for Infrared Spectra Modeling
Authors: Shiluo Huang, Yining Jin, Wei Jin, Ying Mu
AAAI 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirical results show that ACT has achieved competitive results in 9 analytical tasks covering applications across pharmacy, chemistry, and agriculture. Compared with existing networks, ACT reduces the root mean square error of prediction (RMSEP) by more than 20% in calibration transfer tasks. These results indicate that DL methods in IR spectroscopy could benefit from the integration of prior knowledge in analytical chemistry. |
| Researcher Affiliation | Academia | 1 School of Computing and Artificial Intelligence, Southwestern University of Finance and Economics, Chengdu, China 2 Department of Electrical and Computer Engineering, University of Alberta, Edmonton, Alberta, Canada 3 Research Centre for Analytical Instrumentation, State Key Laboratory of Industrial Control Technology, Zhejiang University, Hangzhou, China |
| Pseudocode | No | The paper describes the proposed method using textual descriptions and mathematical equations but does not include any explicit pseudocode blocks or algorithm listings. |
| Open Source Code | Yes | Code https://github.com/Siryokait/ACT4IRSpectra |
| Open Datasets | Yes | The 9 real-world tasks analyzed in this study are derived from 5 open-source datasets. We utilize the standard training, validation, and testing sets provided by these datasets. (1) Tablet dataset (Nikzad-Langerodi et al. 2020) (2) Melamine dataset (Nikzad-Langerodi et al. 2018) (3) Mango DMC dataset (Anderson et al. 2020) (4) Strawberry dataset (Holland, Kemsley, and Wilson 1998) (5) Apple Leaf (Xue et al. 2016) |
| Dataset Splits | Yes | We utilize the standard training, validation, and testing sets provided by these datasets. (1) Tablet dataset ... Tablet(1,2) 155: 40: 460 Regression(CT) Tablet(2,1) 155: 40: 460 Regression(CT) MF(R562,R568) 2122: 910: 733 Regression(CT) MF(R568,R562) 513: 220: 3032 Regression(CT) Tablet(1,1) 155: 40: 460 Regression Tablet(2,2) 155: 40: 460 Regression Mango DMC 7413: 2830: 1448 Regression Strawberry 337: 329: 317 Classification Apple leaf 2500: 1250: 1740 Classification |
| Hardware Specification | Yes | All the experiments are implemented based on Py Torch (Paszke et al. 2019) and are repeated 5 times with NVIDIA RTX 4090 24GB GPU. |
| Software Dependencies | No | All the experiments are implemented based on Py Torch (Paszke et al. 2019). The paper mentions Py Torch but does not specify its version number or any other software dependencies with version numbers. |
| Experiment Setup | Yes | Hyper-parameter settings of ACT is listed in Table 2. The datasets vary in spectral range, resolution, etc., necessitating the tuning of hyper-parameters for each dataset to prevent significant over-fitting or underfitting. In all the tasks, the hyper-parameters of tested methods are adjusted based on data in the calibrating sets. Notably, when dealing with tasks from the same dataset, the hyper-parameters of deep networks remain unchanged. |