AI Classification

The AI classification page serves as a platform for manually adding files as samples to facilitate learning about specific labels. Users can upload files to the AI classification page, providing samples that enable the AI system to better understand and recognize the characteristics associated with particular labels. By adding files manually, users contribute to the training process of the AI model, allowing it to gain insights and improve its accuracy in classifying data according to predefined labels.

AI Classification Page

Users can initiate the process by clicking on upload data which will redirect them to a separate page dedicated to uploading files for each label. Once redirected, users will have the option to upload files specifically associated with individual labels, facilitating targeted learning and training for the AI system.

Adding Files to AI

AI Classification: System Training Options After uploading training files, users can return to the main AI Classification page to select one of three training options for the system:

1. Existing Guard System Files: Utilize the pre-existing training files already available within the Guard system.

2. User-Uploaded Files: Train the system specifically using the files uploaded by the user.

3. Train All Data: Combine both the existing system files and the newly uploaded user files for comprehensive training.

Available Data to Train

Prerequisites for Uploading Files To achieve the best results when training the system, please keep the following recommendations in mind:

• File Quantity: Upload at least 30 files per label to ensure the system has sufficient data for accurate classification.

• Data Quality: The more files you upload, the higher the accuracy of your model will be. Increased file volume allows the AI to better understand and differentiate between labels.

• Balance: Try to upload a similar number of files for each label. Balanced data improves the model’s performance and prevents biases during classification.