MLS-C01 Question 111
Select 3You are working as a Machine Learning Specialist for an e-commerce company. You have created an Amazon SageMaker notebook instance to analyze user purchase data. The dataset contains time-stamped transaction records and the total amount spent per transaction. You want to visualize trends in user spending over time and identify any outliers. Which visualization methods should you use to achieve this?
- A
Scatter plot
- B
Time series plot
- C
Histogram
- D
Box plot
- E
Pie chart
Show answer and explanation
Correct answers: A, B, D
Explanation
To analyze spending trends over time, a time series plot is the most appropriate visualization. A box plot can help identify outliers in spending, which is relevant to this scenario. While scatter plots can show relationships between variables, they are less effective for time-series analysis. Histograms and pie charts are not suitable for this use case as they do not represent trends over time or outliers effectively.
- A. Correct.
Scatter plots are useful for visualizing relationships between two continuous variables. In this scenario, they can show spending trends over time, but they may not capture time-series trends as effectively as a time series plot.
- B. Correct.
Time series plots are specifically designed to show trends over time, making them highly suitable for visualizing user spending trends in this scenario.
- C. Incorrect.
Histograms are used to show the distribution of a single variable, such as transaction amounts, but they do not reveal trends over time or outliers.
- D. Correct.
Box plots are excellent for identifying outliers and understanding the spread of data, such as transaction amounts. They complement time-series analysis well.
- E. Incorrect.
Pie charts are used to show proportions of categories and are not suitable for visualizing trends or outliers in numerical transaction data.