Hyperparameter Tuning

In this lesson, we are going to take a look at how we can improve our models using one of SageMaker's features. In particular, we are going to explore how we can use SageMaker to perform hyperparameter tuning.

In many machine learning models there are some parameters that need to be specified by the model creator and which can't be determined directly from the data itself. Generally the approach to finding the best parameters is to train a bunch of models with different parameters and then choose the model that works best.

SageMaker provides an automated way of doing this. In fact, SageMaker also does this in an intelligent way using Bayesian optimization. What we will do is specify ranges for our hyperparameters. Then, SageMaker will explore different choices within those ranges, increasing the performance of our model over time.

In addition to learning how to use hyperparameter tuning, we will look at Amazon's CloudWatch service. For our purposes, CloudWatch provides a user interface through which we can examine various logs generated during training. This can be especially useful when diagnosing errors.