Sagemaker Processor
Amazon SageMaker allows to build, train, and deploy machine learning models quickly.
Amazon SageMaker is a fully-managed service that covers the entire machine learning workflow to label and prepare your data.
Choose an algorithm, train the algorithm, tune and optimize it for deployment, make predictions, and take action.
Your models get to production faster with much less effort and lower cost.
Once you have trained the model over Sagemaker, you can deploy them as endpoint which can be further utilized in Gathr for prediction over incoming data.
Sagemaker Processor Configuration
To add a SageMaker processor to the pipeline, drag the processor onto the canvas and right-click on it to configure as explained below:
| Field | Description |
|---|---|
| Connection Name | Mention the SageMaker Connection Name. |
| Region | Specify the region name where end point is served. For.e.g., us-west-2. |
| End Point | It lists all the endpoint created in the region mentioned above. |
| Feature Column | All the columns, which serves as input to the model deployed as endpoint over SageMaker. The type of all columns should be double which are mentioned in Feature Column Section. |
| Algorithm Name | Name of the model algorithm, which is getting served as endpoint over Sagemaker. |
| Score | Select the score. |
| Predicted Label | Select the predicted label. |
Out of all built in supported algorithms from Sagemaker, user can utilize below mentioned algorithms endpoints for scoring over incoming data using Gathr.
Linear Learner Binary Classification
Linear Learner Regression
Kmeans
PCA
Tensorflow Classification
XGBoost
Gathr SageMaker supports Custom model as well.
For example, one can train model-using scikit-learn API and then deploy that model as endpoint over SageMaker for prediction.
Once endpoint is in In-service state, it can be used for prediction in Gathr.
The below table has Algorithms with their corresponding output field tabs:
| Field | Description |
|---|---|
| Linear Learner Binary Classification | Score, Predicted Label. |
| Linear Learner Regression | Score. |
| KMeans | ClusterLabel, DistanceToCluster. |
| PCA | Projection. |
| XGBoost | Prediction. |
Tensorflow Classification | Score, OutputClass. |
| Custom | Prediction. |
If you have any feedback on Gathr documentation, please email us!