Skip to main content
Fig. 6 | BMC Bioinformatics

Fig. 6

From: Using transfer learning and dimensionality reduction techniques to improve generalisability of machine-learning predictions of mosquito ages from mid-infrared spectra

Fig. 6

MLP trained on PCA-transformed Ifakara dataset plus 2% new target population samples: A As training time increased (5000 epochs), training and validation classification accuracy for mosquito age classes increased from 50 to 91%, B A normalised confusion matrix displaying the proportions of correct mosquito age class predictions achieved on the held-out Ifakara test set during model training, C Proportions of correct mosquito age class predictions achieved on unseen Glasgow dataset. MLP trained on t-SNE-transformed Ifakara dataset plus 2% new target population samples: D As training time increased (5000 epochs), training and validation classification accuracy for mosquito age classes increased from 60 to 83%, E A normalised confusion matrix displaying the proportions of correct mosquito age class predictions achieved on the held-out Ifakara test set during model training, F Proportions of correct mosquito age class predictions achieved on unseen Glasgow dataset

Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us