How to train a model from MariaDB?
How to train a new model
To train a new model, you will need to
INSERT a new record inside the mindsdb.predictors table.
How to create the mindsb.predictors table
Note that after connecting MindsDB and MariaDB, on start, the MindsDB server will automatically create the mindsdb database and add the predictors table.
Don't forget to enable CONNECT Storage Engine as explained in connect your data section.
INSERT query for training new model is quite simple, e.g.:
INSERT INTO mindsdb.predictors(name, predict, select_data_query, training_options) VALUES('model_name', 'target_variable', 'SELECT * FROM table_name');
- name (string) -- The name of the model.
- predict (string) -- The feature you want to predict. To predict multiple features, include a comma separated string, e.g. 'feature1,feature2'.
- select_data_query (string) -- The SELECT query that will ingest the data to train the model.
- training_options (JSON as comma separated string) -- optional value that contains additional training parameters. For a full list of parameters, check the PredictorInterface.
To train timeseries model, check out the timeseries example.
Train new model example
The following example shows you how to train a new model from the MariaDB client. The table used for training the model is the Used cars dataset.
INSERT INTO mindsdb.predictors(name, predict, select_data_query) VALUES ('used_cars_model', 'price', 'SELECT * FROM data.used_cars_data');
INSERT query will train a new model called
used_cars_model that predicts the cars
How to check model training status?
To check that the training finished successfully, you can
SELECT from the mindsdb.predictors table and get the training status, e.g.:
SELECT * FROM mindsdb.predictors WHERE name='<model_name>';
You have successfully trained a new model from the MariaDB database. The next step is to get predictions by querying the model.