Effect of machine learning techniques to detect Listeria monocytogenes in Queso fresco using shortwave-infrared imaging
Bibliographic record
Abstract
Queso fresco (QF) is a type of soft, fresh cheese, often prone to post-processing Listeria monocytogenes (LM) contamination. In this study, we evaluated the potential of shortwave infrared (SWIR) imaging to detect LM in QF. About 10 g of QF was surface inoculated with three different strains of LM, such that the final population was approximately 1.0 log 10 CFU/g, 2.0 log 10 CFU/g, and 3.0 log 10 CFU/g. Following image acquisition, statistical features namely mean reflectance, standard deviation of reflectance, skewness, and kurtosis were used to develop classification models. A trend of decrease in mean reflectance with increase in LM population was observed. Three types of classification (binary, population-wise, and population-strain-wise) were performed by four supervised machine learning (ML) algorithms - Logistic regression (LR), Random Forest (RF), Support vector machine (SVM), and k-Nearest neighbor (kNN). RF outperformed binary and population-wise classifications with an accuracy of 100 %. In binary classification, followed by RF, SVM and kNN exhibited an accuracy of 94 % and 92 % respectively. In population-wise classification, SVM and kNN had classification accuracies in the range of 85–88 %. Among the ML models, LR resulted in poor accuracies across all three classifications. Strain-wise classification did not yield reliable accuracies, implying the overlap in genetic similarities. This study demonstrates that SWIR imaging along with chemometrics can be a prospective tool for real-time detection and (or) quantification of LM in fresh cheeses like QF. This approach will likely be a novel safety assessment tool in cheese industry with the potential to enhance product safety and consumer confidence in consumption of fresh cheeses.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".