- Categorical Data
- Training data that assigns a text value to individual examples. Query results try to find the label (category) that the query term most closely resembles. For instance, the Hello World example tries to determine whether the string passed in resembles entries labeled as "English," "Spanish," or "French" most closely.
- Another name for using the categorical data system.
- Continuous Data
- Another term for regression data.
- A single entry in the training data. For example, in spam detection training data, each email is a single example.
- Example value
- The value assigned to a single example in the training data. This is the first column in the training data, and is numeric for regression models and text for categorical models.
- An individual data value within a training (or query) entry. For example, if a training entry for user height consists of person's height, person's gender, mother's height, father's height, and country of origin, each of those elements is called a feature.
- Hosted model
- A pre-trained model in our gallery. Hosted models are useful when you don't have the time, data, or resources to develop your own model for a specific use.
- Streaming training
- Sending additional training examples to an already trained model.
- [Trained] Model
- When you train the Prediction API against a data set, it creates a model based on that specific training data. All queries for the Prediction API are sent to a model trained against a specific training data.
- Regression Data
- Training data that assigns numeric values to individual examples. Regression models enable the API to predict numeric values based on examples. For example, to return a predicted temperature, based on prior temperature entries.
- See Streaming training
Monitor your resources on the go
Get the Google Cloud Console app to help you manage your projects.