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

Understanding why a neural network model makes certain decisions can be asimportant as the inference performance. Various methods have been proposed tohelp practitioners explain the prediction of a neural network model, of whichShapley values are most popular. SHAP package is a leading implementation ofShapley values to explain neural networks implemented in TensorFlow or PyTorchbut lacks cross-platform support, one-shot deployment and is highlyinefficient. To address these problems, we present the ONNXExplainer, which isa generic framework to explain neural networks using Shapley values in the ONNXecosystem. In ONNXExplainer, we develop its own automatic differentiation andoptimization approach, which not only enables One-Shot Deployment of neuralnetworks inference and explanations, but also significantly improves theefficiency to compute explanation with less memory consumption. For faircomparison purposes, we also implement the same optimization in TensorFlow andPyTorch and measure its performance against the current state of the artopen-source counterpart, SHAP. Extensive benchmarks demonstrate that theproposed optimization approach improves the explanation latency of VGG19,ResNet50, DenseNet201, and EfficientNetB0 by as much as 500%.

ONNXExplainer: an ONNX Based Generic Framework to Explain Neural Networks Using Shapley Values