Data Annotation (labels)
t6 is focused on timeseries, Data-annotation process is classifying Datapoints from Flows using categories. Annotations and Categories are going to be used in the Exploratory Data Analysis process and in Machine-Learning.
Tagged on
#data-annotation,
#label,
#labellisation,
#feature,
Categories
A classification category in t6 consists of a name, descriptive content, and a color. Users have the flexibility to customize their own categories, although there are constraints on the number of categories per user. There is no approval process for annotations, allowing users to freely define and utilize categories as needed. These categories are essential for machine learning tasks, as they serve as the basis for training models.
Annotations / labels
Data annotation can be associated with datapoints in t6 using two primary methods:
- Manual Annotation: Users can manually add annotations to datapoints using the specific annotation API endpoint. This method offers strong supervision and precision but requires manual intervention.
- Programmatic Annotation: Annotations can also be added programmatically from a decision rule during datapoint creation. This approach allows for automated annotation based on predefined criteria within the decision rules. Annotations can be set for specific date/time ranges, enabling users to define category associations within a specific timeframe. Both manual and programmatic annotation processes are designed to handle large datasets efficiently. By leveraging categories and annotations, users can effectively organize and classify their data, paving the way for advanced data analysis and machine learning applications.