Fine-Tune the Regression Model

In this example, we use our sample PostgreSQL database. You can connect to it like this:

CREATE DATABASE example_db
WITH ENGINE = "postgres",
PARAMETERS = {
    "user": "demo_user",
    "password": "demo_password",
    "host": "samples.mindsdb.com",
    "port": "5432",
    "database": "demo"
    };

First, we create and train the model using a subset of the home_rentals data, considering properties that have been on the market less than 10 days.

CREATE MODEL mindsdb.adjust_home_rentals_model
FROM example_db
    (SELECT * 
    FROM demo_data.home_rentals 
    WHERE days_on_market < 10)
PREDICT rental_price;

On execution, we get:

Query successfully completed

We can check its status using this command:

SELECT *
FROM mindsdb.models
WHERE name = 'adjust_home_rentals_model';

Once the status is complete, we can query for predictions.

SELECT rental_price, rental_price_explain 
FROM mindsdb.adjust_home_rentals_model
WHERE sqft = 1000
AND location = 'great'
AND neighborhood = 'berkeley_hills'
AND number_of_rooms = 2
AND number_of_bathrooms = 1
AND days_on_market = 40;

On execution, we get:

+---------------+-----------------------------------------------------------------------------------------------------------------------------------------------+
| rental_price  | rental_price_explain                                                                                                                          |
+---------------+-----------------------------------------------------------------------------------------------------------------------------------------------+
| 2621          | {"predicted_value": 2621, "confidence": 0.99, "anomaly": null, "truth": null, "confidence_lower_bound": 2523, "confidence_upper_bound": 2719} |
+---------------+-----------------------------------------------------------------------------------------------------------------------------------------------+

Let’s adjust this model with more training data. Now we consider properties that have been on the market for 10 or more days.

FINETUNE mindsdb.adjust_home_rentals_model
FROM example_db
    (SELECT * 
    FROM demo_data.home_rentals 
    WHERE days_on_market >= 10);

While the model is being generated and trained, it is not active. The model becomes active only after it completes generating and training.

To check the status and versions of the model, run this command:

SELECT name, engine, project, active, version, status
FROM mindsdb.models_versions
WHERE name = 'adjust_home_rentals_model';

On execution, we get:

+---------------------------+-----------+---------+--------+---------+----------+
| name                      | engine    | project | active | version | status   |
+---------------------------+-----------+---------+--------+---------+----------+
| adjust_home_rentals_model | lightwood | mindsdb | false  | 1       | complete |
| adjust_home_rentals_model | lightwood | mindsdb | true   | 2       | complete |
+---------------------------+-----------+---------+--------+---------+----------+

Please note that the longer the property is on the market, the lower its rental price. Hence, we can expect the rental_price prediction to be lower.

SELECT rental_price, rental_price_explain 
FROM mindsdb.adjust_home_rentals_model
WHERE sqft = 1000
AND location = 'great'
AND neighborhood = 'berkeley_hills'
AND number_of_rooms = 2
AND number_of_bathrooms = 1
AND days_on_market = 40;

On execution, we get:

+---------------+-----------------------------------------------------------------------------------------------------------------------------------------------+
| rental_price  | rental_price_explain                                                                                                                          |
+---------------+-----------------------------------------------------------------------------------------------------------------------------------------------+
| 2055          | {"predicted_value": 2055, "confidence": 0.99, "anomaly": null, "truth": null, "confidence_lower_bound": 1957, "confidence_upper_bound": 2153} |
+---------------+-----------------------------------------------------------------------------------------------------------------------------------------------+

If you have dynamic data that gets updated regularly, you can set up an automated fine-tuning as below.

Note that the data source must contain an incremental column, such as timestamp or integer, so MindsDB can pick up only the recently added data.

Create a view to store recently added data with the help of the LAST keyword:

CREATE VIEW recent_data (
    SELECT *
    FROM example_db.home_rentals
    WHERE timestamp > LAST
);

Create a job to fine-tune the model periodically.

CREATE JOB automated_finetuning (

    FINETUNE adjust_home_rentals_model
    FROM mindsdb
        (SELECT * FROM recent_data)
)
EVERY 1 day;

Now your model will be fine-tuned with newly added data every day.

Use Cases