ML Pipeline Day 2

You have been tasked with transforming data that is stored in Amazon Relational Database Service (RDS) into Amazon S3. Currently you have Multi-AZ RDS enabled and setup inside a private VPC subnet with no access to the outside internet. You have setup an AWS Glue job to run using JDBC connection running in the same private VPC. Which of the following steps will occur or be applied to enable your transformation job to run successfully and securely? (SELECT TWO)

• Setup Internet Gateway attached to private VPC blocking all outside connections

• Setup VPC peering for the RDS instance inside the private VPC

• Setup a Network Address Translation (NAT) gateway inside the VPC.

• Setup a VPC Gateway Endpoint to access S3 as your data destination

• AWS Glue sets up elastic network interfaces that enable your jobs to connect securely to RDS within your VPC

You are a machine learning specialist evaluating a current model that has been deployed into production. It has been deployed for a few weeks now and the results are not accurate and sometimes the inference data is missing values. What are some techniques you can review to help solve this problem? (SELECT THREE)

• Ensure the target variable used as the predictor during training represents the actual outcome that the machine learning model is trying to predict.

• Ensure the training data has a 50/50 distribution of the target attribute.

• Ensure the inference data has placeholder values for any of the missing values

• Ensure the training datasets are large, representative samples of the populations that the model needs to make predictions.

• Ensure the extraction methods used to generate the training datasets are the same as for the production inference data.

You have been tasked with creating a labeled dataset by classifying text data into different categories depending on the summary of the corpus. You plan to use this data with a particular machine learning algorithm within AWS. Your goal is to make this as streamlined as possible with minimal amount of setup from you and your team. What tool can be used to help label your dataset with the minimum amount of setup?

• Amazon Latent Dirichlet Allocation (LDA) algorithm

• Marketplace AMI for NLP problems

• AWS Comprehend entity detection

• Amazon Neural Topic Modeling (NTM) built-in algorithm

• AWS SageMaker GroundTruth text classification job

You are a machine learning specialist finding ways to detect anomalous data points within a given labeled data set. You've been tasked with creating a model to achieve this and also determine how accurate the model is along with other metrics like precision, recall, and F1-score metrics on the labeled data. How can this easily be achieved?

• Create a model using the Random Cut Forest (RCF) algorithm with both a train and the optional test data channels. Use application/json for training and validation data. Train the model on an ml.m4 or ml.c4 instance type.

• Create a model using the XGBoost algorithm with both a train and optional validation channels. Use application/x-recordio-protobuf for training and validation data. Train the model on an ml.c4 or ml.g4 instance type.

• Create a model using the XGBoost algorithm with both a train and the optional test data channels. Use application/x-recordio-protobuf for training and validation data. Train the model on an ml.m4 or ml.g4 instance type.

• Create a model using the Random Cut Forest (RCF) algorithm with both a train and the optional test data channels. Use text/csv for training and validation data. Train the model on an ml.m4 or ml.c4 instance type.

You are working with a colleague in your company on creating some documentation for using SageMaker internally with XGBoost. Your colleague is based in France while you are based in Canada. You are reviewing her documentation and notice that the training image registry path does not match the path that you have recorded in your version. What is the most likely reason for this?

• Built-in algorithm registry paths are randomly generated for each account.

• You both are using different regions.

• You selected a training image while your colleague selected an inference image.

• Your colleague has selected the wrong version of XGBoost.