Deep learning model insights using provenance data
A method, computer system, and a computer program product for generating deep learning model insights using provenance data is provided. Embodiments of the present invention may include collecting provenance data. Embodiments of the present invention may include generating model insights based on the collected provenance data. Embodiments of the present invention may include generating a training model based on the generated model insights. Embodiments of the present invention may include reducing the training model size. Embodiments of the present invention may include creating a final trained model.
1. A method comprising:
collecting provenance data;
generating model insights based on the collected provenance data;
generating a training model based on the generated model insights;
reducing the training model size; and
creating a final trained model;
classifying a query sample using the final trained model;
determining a misclassification of the query sample in a final trained model output;
analyzing the misclassification using the provenance data; and
providing generated insights based on the misclassification to a user.
2. The method of claim 1 , wherein the provenance data includes data collected during a deep learning training phase.
3. The method of claim 1 , wherein the provenance data includes learned weights, learned activation values, learned nodes and learned gradients based on a training image being passed through one or more iterations of a deep learning model.
4. The method of claim 1 , wherein the provenance data includes learned weights, learned activation values, learned nodes and learned gradients based on a training image being passed through one or more layers of a deep learning model.
5. The method of claim 1 , wherein the model insights are generated by analyzing the provenance data, wherein the model insights include an analysis of neuron weights, neuron learning features, neuron histories, neuron lineages, correlated neurons and unused neurons.
6. The method of claim 1 , wherein reducing the model size is determined based on a periodic analysis.
7. A computer system comprising:
one or more processors, one or more computer-readable memories, one or more computer-readable tangible storage media, and program instructions stored on at least one of the one or more computer-readable tangible storage media for execution by at least one of the one or more processors via at least one of the one or more computer-readable memories, wherein the computer system is capable of performing a method comprising:
collecting provenance data;
generating model insights based on the collected provenance data;
generating a training model based on the generated model insights;
reducing the training model size; and
creating a final trained model;
classifying a query sample using the final trained model;
determining a misclassification of the query sample in a final trained model output;
analyzing the misclassification using the provenance data; and
providing generated insights based on the misclassification to a user.
8. The computer system of claim 7 , wherein the provenance data includes data collected during a deep learning training phase.
9. The computer system of claim 7 , wherein the provenance data includes learned weights, learned activation values, learned nodes and learned gradients based on a training image being passed through one or more iterations of a deep learning model.
10. The computer system of claim 7 , wherein the provenance data includes learned weights, learned activation values, learned nodes and learned gradients based on a training image being passed through one or more layers of a deep learning model.
11. The computer system of claim 7 , wherein the model insights are generated by analyzing the provenance data, wherein the model insights include an analysis of neuron weights, neuron learning features, neuron histories, neuron lineages, correlated neurons and unused neurons.
12. The computer system of claim 7 , wherein reducing the model size is determined based on a periodic analysis.
13. A computer program product comprising:
one or more computer-readable tangible storage media and program instructions stored on at least one of the one or more computer-readable tangible storage media, the program instructions executable by a processor to cause the processor to perform a method comprising:
collecting provenance data;
generating model insights based on the collected provenance data;
generating a training model based on the generated model insights;
reducing the training model size; and
creating a final trained model;
classifying a query sample using the final trained model;
determining a misclassification of the query sample in a final trained model output;
analyzing the misclassification using the provenance data; and
providing generated insights based on the misclassification to a user.
14. The computer program product of claim 13 , wherein the provenance data includes data collected during a deep learning training phase.
15. The computer program product of claim 13 , wherein the provenance data includes learned weights, learned activation values, learned nodes and learned gradients based on a training image being passed through one or more iterations of a deep learning model.
16. The computer program product of claim 13 , wherein the provenance data includes learned weights, learned activation values, learned nodes and learned gradients based on a training image being passed through one or more layers of a deep learning model.
17. The computer program product of claim 13 , wherein the model insights are generated by analyzing the provenance data, wherein the model insights include an analysis of neuron weights, neuron learning features, neuron histories, neuron lineages, correlated neurons and unused neurons.