What to Do If Machine Learning Classification Fails to Achieve Expected Results?

• The effectiveness of machine learning classification depends on the matching degree between training samples and testing samples. Ensure training sample labels are accurate and maintain high overlap with testing samples. Multiple approaches should be experimented with, focusing on datasets that demonstrate better classification performance for further optimization.

  Is It Possible to Delete a Specific Class of Points, Such as Low Points?

• LiDAR360 supports extracting point cloud data by class. Navigate to the "Tools" menu, select the "Extract" dropdown, and choose "Extract by Class." Initially, deselect the category you wish to remove (e.g., "Low Points") to exclude the corresponding data.