Salesforce, inc. (20240241820). SYSTEMS AND METHODS FOR PROVIDING AN AUTOMATED TESTING PIPELINE FOR NEURAL NETWORK MODELS simplified abstract
Contents
SYSTEMS AND METHODS FOR PROVIDING AN AUTOMATED TESTING PIPELINE FOR NEURAL NETWORK MODELS
Organization Name
Inventor(s)
Shiva Kumar Pentyala of Mountain View CA (US)
Shashank Harinath of Mountain View CA (US)
Sitaram Asur of Newark CA (US)
Zachary Alexander of Berkeley CA (US)
SYSTEMS AND METHODS FOR PROVIDING AN AUTOMATED TESTING PIPELINE FOR NEURAL NETWORK MODELS - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240241820 titled 'SYSTEMS AND METHODS FOR PROVIDING AN AUTOMATED TESTING PIPELINE FOR NEURAL NETWORK MODELS
The abstract describes an automated testing pipeline for creating a testing dataset to test a trained neural network model. The pipeline involves receiving a first testing dataset with user queries, filtering queries based on action verbs, ranking queries using a pretrained language model, removing keyword matches with a bag of words classifier, and generating a second testing dataset for the neural network model to produce testing outputs.
- Automated testing pipeline for testing neural network models
- Filtering user queries based on action verbs
- Ranking queries using a pretrained language model
- Removing keyword matches with a bag of words classifier
- Generating a second testing dataset for testing outputs
Potential Applications: - Quality assurance in machine learning models - Improving accuracy and reliability of neural network testing - Enhancing performance of trained models in real-world applications
Problems Solved: - Streamlining the testing process for neural network models - Ensuring the accuracy and relevance of testing datasets - Enhancing the overall performance of trained models
Benefits: - Increased efficiency in testing neural network models - Improved accuracy and reliability of testing results - Enhanced performance of trained models in real-world scenarios
Commercial Applications: Title: Automated Testing Pipeline for Neural Network Models This technology can be utilized in industries such as: - Software development - Artificial intelligence research - Data analytics companies
Questions about Automated Testing Pipeline for Neural Network Models:
1. How does the automated testing pipeline improve the efficiency of testing neural network models? The automated testing pipeline streamlines the process of creating testing datasets and ensures the accuracy and relevance of the data used for testing, ultimately leading to improved efficiency in testing neural network models.
2. What are the key features of the automated testing pipeline for neural network models? The key features include filtering user queries based on action verbs, ranking queries using a pretrained language model, removing keyword matches with a bag of words classifier, and generating a second testing dataset for testing outputs.
Original Abstract Submitted
embodiments described herein provide an automated testing pipeline for providing a testing dataset for testing a trained neural network model trained using a first training dataset. a first testing dataset for the trained neural network including a first plurality of user queries is received. a dependency parser is used to filter the first plurality of user queries based on one or more action verbs. a pretrained language model is used to rank the remaining user queries based on respective relationships with queries in the first training dataset. further, user queries that are classified as keyword matches with the queries in the first training dataset using a bag of words classifier are removed. a second testing dataset is generated using the ranked remaining user queries. testing outputs are generated, by the trained neural network model, using the second testing dataset.
