Model Performance Monitoring

How do you monitor model metrics - Mean squared, R1, TP/TN/FP/FN ? Tools/Solutions?

Hello, Kaviajay!

These are the metrics available in the TensorFlow Keras API:

The simplest way to monitor them is to select your metrics as a list, and add them to the argument metrics in model.compile. An example:

model.compile(
    optimizer='adam',
    loss='mean_squared_error',
    metrics=[
        'MeanSquaredError',
        'AUC',
    ]
)

You can read more about how to use metrics here:

We’ll be glad to answer any more questions that remain!