Improving Predictions through feedback
Active learning is the process of providing feedback to your user-created enrichment models to help improve the model’s performance. By reviewing generated predictions and confirming their accuracy or updating them to a more appropriate value, a user is able to provide constructive feedback, retrain the model and generate refined predictions for your entities.
Graft helps you optimize your feedback by providing a prioritized list of those predictions we feel will provide the most impact to your model, reducing the time taken to tune the model. This feedback can be bite size with incremental changes applied as needed, with no requirement to manually label significant portions of your data set at any one time.
Active Learning within Graft allows you to create, manage and refine your models and simplifies your data management processes removing the need to spend time on long, manually intensive, processes to update your data in external systems before importing and reprocessing to generate new models.
Start enhancing your models by following our Active Learning guide