English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep convolutional neural networks have proven their effectiveness, and havebeen acknowledged as the most dominant method for image classification.However, a severe drawback of deep convolutional neural networks is poorexplainability. Unfortunately, in many real-world applications, users need tounderstand the rationale behind the predictions of deep convolutional neuralnetworks when determining whether they should trust the predictions or not. Toresolve this issue, a novel genetic algorithm-based method is proposed for thefirst time to automatically evolve local explanations that can assist users toassess the rationality of the predictions. Furthermore, the proposed method ismodel-agnostic, i.e., it can be utilised to explain any deep convolutionalneural network models. In the experiments, ResNet is used as an example modelto be explained, and the ImageNet dataset is selected as the benchmark dataset.DenseNet and MobileNet are further explained to demonstrate the model-agnosticcharacteristic of the proposed method. The evolved local explanations on fourimages, randomly selected from ImageNet, are presented, which show that theevolved local explanations are straightforward to be recognised by humans.Moreover, the evolved explanations can explain the predictions of deepconvolutional neural networks on all four images very well by successfullycapturing meaningful interpretable features of the sample images. Furtheranalysis based on the 30 runs of the experiments exhibits that the evolvedlocal explanations can also improve the probabilities/confidences of the deepconvolutional neural network models in making the predictions. The proposedmethod can obtain local explanations within one minute, which is more than tentimes faster than LIME (the state-of-the-art method).

Explaining Deep Convolutional Neural Networks for Image Classification by Evolving Local Interpretable Model-agnostic Explanations