As with any machine learning problem, the quality of the predictions depends heavily on the quality of the data. Although we have a robust data cleaning algorithm, you may find some predictions to be highly inaccurate, while others might be predicted with 100% accuracy.

This discrepancy could be due to the fact that the data provided as input was insufficient for our machine learning pipelines to run efficiently. Rest assured, your data is secure with us. Please read our security policy and consider providing a larger dataset of defects for improved results.

You may want to try the pipeline again with a fresh new data, Try Now!