AWS Certified Data Analytics Specialty 2021 – Hands-On - (Exercise) Amazon Machine Learning, Part 2

The video tutorial covers the process of training and evaluating a regression model using Amazon Machine Learning. It explains the difference between numerical regression and classification problems, and highlights the importance of data cleaning to improve model accuracy. The tutorial demonstrates real-time predictions and discusses the evaluation results, emphasizing the need for data preparation. It concludes with a cleanup of the model to avoid unnecessary charges.

What is the main challenge mentioned in predicting quantities based on past behavior?

• Limited data availability

• Wide range of quantities ordered

• Lack of suitable algorithms

• High computational cost

What type of problem is it when predicting a number with meaningful order?

• Binary classification

• Multiclass classification

• Clustering

• Numerical regression

Which of the following is NOT a type of classification problem?

• Binary classification

• Numerical regression

• Hierarchical classification

• Multiclass classification

What was the model's quality score compared to the baseline?

• The same

• Much better

• Slightly better

• Significantly worse

What was the predicted quantity for the first row of sample data?

• 14

• 10

• 6

• 20

Why is data cleaning important in machine learning?

• To simplify algorithms

• To improve model accuracy

• To reduce computational time

• To increase data size

What was identified as a major issue in the dataset used for the model?

• Outliers

• Duplicate entries

• Missing values

• Incorrect labels