ML Pipeline

A financial planning company is using the Amazon SageMaker endpoint with an Auto Scaling policy to serve its forecasting model to the company’s customers to help them plan for retirement. The team wants to update the endpoint with its latest forecasting model, which has been trained using Amazon SageMaker training jobs. The team wants to do this without any downtime and with minimal change to the code.

• Use a new endpoint configuration with the latest model S3 path in the UpdateEndpoint API

• Update the endpoint using a new configuration with the lastest model S3 path. Then, register the endpoint as a scalable target

• Create a new endpoint using a new configuration with the latest model. Then register the endpoint as a scalable target.

• De-register the endpoint as a scalable target. Update the endpoint using a new endpoint configuration with the latest S3 path. Finally, register the endpoint as scalable target again.

A navigation and transportation company is using satellite images to model weather around the world in order to create optimal routes for its ships and planes. The company is using Amazon SageMaker training jobs to build and train its models.

However, during training, it takes too long to download the company’s 100 GB data from Amazon S3 to the training instance before the training starts.

What should the company do to speed up its training jobs while keeping the costs low?

• Change the input to Pipe mode

• Increase the batch size in the model

• Increase the instance size for training

• Create an EBS volume with the data on it and attach it to the training job

A real estate startup wants to use ML to predict the value of homes in various cities. To do so, the startup’s data science team is joining real estate price data with other variables such as weather, demographic, and standard of living data.

However, the team is having problems with slow model convergence. Additionally, the model includes large weights for some features, which is causing degradation in model performance.

What kind of data preprocessing technique should the team use to more effectively prepare this data?

• Max absolute scaler

• One Hot encoder

• Normalizer

• Standard Scaler

A multi-national banking organization provides loan services to customers worldwide. Many of its customers still submit loan applications in paper form in one of the bank’s branch locations. The bank wants to speed up the loan approval process for this set of customers by using machine learning. More specifically, it wants to create a process in which customers submit the application to the clerk, who scans and uploads it to the system. The system then reads and provides an approval or denial of the application in a matter of minutes.

What can the bank use to read and extract the necessary data from the loan applications without needing to manage the process?

• A custom CNN model

• A LSTM model

• Amazon Personalize

• Amazon Textract

A ride-share company wants to create intelligent conversational chatbots that will serve as first responders to customers who call to report an issue with their ride. The company wants these chatbot-customer calls to mimic natural conversations that provide personalized experiences for the customers.

What combination of AWS services can the company use to create this workflow without a lot of ongoing management?

• Amazon Transcribe to parse the utterances and intent of customers comments, Amazon Lex to reply to the customers

• Amazon Transcribe to parse the utterances and intent of customers comments, Amazon Polly to reply to the customers

• Amazon Lex to parse the utterances and intent of customers comments, Amazon Polly to reply to the customers

• Amazon Polly to parse the utterances and intent of customers comments, Amazon Lex to reply to the customers