Main Content

Visualization and Verification

Visualize neural network behavior, explain predictions, and verify robustness using sequence and tabular data

Visualize deep networks during and after training. Monitor training progress using built-in plots of network accuracy and loss. To investigate trained networks, you can use visualization techniques such as Grad-CAM.

Use deep learning verification methods to assess the properties of deep neural networks. For example, you can verify the robustness properties of a network, compute network output bounds, and find adversarial examples.


Deep Network DesignerDesign and visualize deep learning networks


expand all

analyzeNetworkAnalyze deep learning network architecture
trainingProgressMonitorMonitor and plot training progress for deep learning custom training loops (Since R2022b)
updateInfoUpdate information values for custom training loops (Since R2022b)
recordMetricsRecord metric values for custom training loops (Since R2022b)
groupSubPlotGroup metrics in training plot (Since R2022b)
plotPlot neural network architecture
predictCompute deep learning network output for inference (Since R2019b)
minibatchpredictMini-batched neural network prediction (Since R2024a)
scores2labelConvert prediction scores to labels (Since R2024a)
gradCAMExplain network predictions using Grad-CAM (Since R2021a)
confusionchartCreate confusion matrix chart for classification problem
sortClassesSort classes of confusion matrix chart
rocmetricsReceiver operating characteristic (ROC) curve and performance metrics for binary and multiclass classifiers (Since R2022b)
addMetricsCompute additional classification performance metrics (Since R2022b)
averageCompute performance metrics for average receiver operating characteristic (ROC) curve in multiclass problem (Since R2022b)


ConfusionMatrixChart PropertiesConfusion matrix chart appearance and behavior
ROCCurve PropertiesReceiver operating characteristic (ROC) curve appearance and behavior (Since R2022b)



Training Progress and Performance