Knowledge Check

The following questions will help you check what you've learned.

1. Question 1 of 6What is the goal of feature engineering in the machine learning lifecycle?

   ATo collect and clean dataBTo transform data and create variables for the modelCTo calculate the model performanceDTo maintain the desired level of performance
   Check answer
2. Question 2 of 6A small e-commerce company wants to use machine learning to improve their churn prediction. However, the company does not have a dedicated data science team and is looking for a low-code or no-code solution to get started with machine learning. Which Amazon SageMaker AI feature would be the most suitable for the company's requirements?

   AAmazon SageMaker CanvasBAmazon SageMaker JumpStartCAmazon SageMaker Data WranglerDAmazon SageMaker Feature Store
   Check answer
3. Question 3 of 6A company wants to implement a proof of concept about price predictions for a set of items. They are starting from scratch and want a solution with minimal operational overhead. Which implementation should the company use?

   AAmazon SageMaker JumpStartBA custom docker imageCA custom script using a supported third-party frameworkDA custom script using a built-in SageMaker AI algorithm
   Check answer
4. Question 4 of 6A company develops a machine learning model to predict the price of a house based on its features, such as the number of bedrooms, bathrooms, square footage, and location. The model is trained on a dataset of historical house sales, and its performance is evaluated using various metrics. However, the model is found to perform well on the training data but poorly on new, unseen data. Which option describes the fit of model?

   AIt overfits.BIt underfits.CIt has a low variance.DIt has a high bias.
   Check answer
5. Question 5 of 6A real-time object detection model has been developed by a team of developers for a self-driving car company to identify various objects in the environment for their autonomous vehicle. The model must be deployed in a way that ensures minimal latency and must be capable of interacting with the vehicle's onboard systems. Which AWS deployment options should the team choose to deploy the model to achieve minimal latency while enabling real-time interactions with the onboard systems?

   AReal-timeBAsynchronousCBatchDServerless
   Check answer
6. Question 6 of 6Which option best describes machine learning operations (MLOps)?

   AMLOps is a framework that focuses solely on the development and training of ML models.BMLOps is a process that involves deploying ML models in a production environment and maintaining them over time.CMLOps is a set of practices that aims to manage the end-to-end lifecycle of machine learning (ML) systems, including development, deployment, monitoring, and maintenance.DMLOps is a tool used for visualizing and analyzing the performance of machine learning (ML) models.
   Check answer

You learned about the machine learning lifecycle, how to leverage AWS service at every step. You also learned techniques to evaluate ML models performance, and understand the importance of MLOps in streamlining the development and deployment of your machine learning projects.