You are building a production dashboard to monitor a newly deployed recommendation model. You need to display both the model's prediction accuracy (measured as a percentage) and its inference latency (measured in milliseconds) over a 30-day period on a single chart to help the team spot correlations, such as accuracy drops during high-latency periods. Which visualization style is best suited for this task?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Now, if you look at this from a troubleshooting perspective, you want to see if your model gets dumber when it gets slower. But think of it: accuracy is a percentage (like 98%), and latency is in milliseconds (like 50ms). If you put those on the same single axis, your accuracy line will look like a flat ceiling and your latency will be all over the place — or vice versa. It's useless. The clean way to do this is with a dual-axis line chart. Time goes on the bottom (X-axis), accuracy goes on the left (Y1), and latency goes on the right (Y2). Now you can see both trends side-by-side, clear as day. Got it? Sweet. Let's keep rolling.
Full explanation below image
Full Explanation
Monitoring models in production requires visualizing multiple metrics to identify performance degradation, such as model drift or resource bottlenecks. Accuracy (ranging from 0% to 100%) and latency (measured in milliseconds or seconds) have completely different scales and units. A dual-axis line chart (Option B) is the ideal visualization for this scenario because it plots two distinct metrics against a shared independent variable (time, on the X-axis) using two separate vertical axes (Y-axes). This allows users to easily view trends, detect anomalies, and identify temporal correlations (such as latency spikes coinciding with accuracy drops) without scale distortion. A box plot (Option A) is useful for showing data distribution and outliers but discards the chronological sequence, making it impossible to track trends over time. A stacked area chart (Option C) adds the values together, which is mathematically nonsensical for percentage accuracy and millisecond latency. A pie chart (Option D) shows parts of a whole at a single point in time, which cannot display temporal trends. Therefore, a dual-axis line chart is the most effective choice.