in

Neuron classification using Sugeno fuzzy integration and multi-classifier fusion #NeuronClassification

Morphological classification of neurons based on Sugeno fuzzy integration and multi-classifier fusion

The content discusses the evaluation of a model’s classification performance using various criteria such as confusion matrix, ROC curve, and AUC region, as well as accuracy, precision, recall rate, and F1 value. The experiment involves classifying neuron images from the NeuroMorpho.Org dataset into different categories. The results show that an improved model, MCF-Net, outperforms other models in accuracy for both 4-category and 12-category classifications. The model fusion using Sugeno fuzzy integral algorithm achieves the highest classification accuracy. The computational complexity analysis reveals that MCF-Net has a large number of parameters but achieves high accuracy. GradCAM analysis shows different models focusing on different regions of neuron images, highlighting the complementarity of models and the significance of model fusion. Comparison with existing methods demonstrates the superiority of the proposed model in neuron morphology classification. Overall, the study showcases the effectiveness of model fusion in improving classification accuracy and the importance of considering computational complexity in model selection.

Source link

Source link: https://www.nature.com/articles/s41598-024-66797-1

What do you think?

Leave a Reply

GIPHY App Key not set. Please check settings

AI in 2024 Pt. 2: An Astrological Timeline of Key Developments for the Next 6 Months

Astrological timeline of AI developments in 2024 with #predictions

Empowering NGOs with generative AI in the fight against human trafficking

AI empowering NGOs to combat human trafficking with #generativeAI.