Training and deploying an RCF model
In this recipe, we will train and deploy an RCF model using the SageMaker Python SDK. The RCF algorithm is an unsupervised algorithm used for detecting anomalies in a dataset. It associates each record with an anomaly score value, with higher anomaly score values associated with records that may potentially be tagged as outliers or anomalies.
After we have trained and deployed an RCF model in this recipe, we will trigger this model within an Amazon Athena SQL query in the Invoking machine learning models with Amazon Athena using SQL queries recipe. This will enable us to tag anomalies in our dataset during the data preparation and analysis phase.
Getting ready…
This recipe continues from Generating a synthetic dataset for anomaly detection experiments.
How to do it…
The next set of steps focus on using the dataset we generated in the previous recipe to prepare the RCF model:
- Navigate to the
my-experiments/chapter04...
