Free Amazon MLS-C01 Actual Exam Questions - Question 14 Discussion
A machine learning (ML) specialist at a retail company must build a system to forecast the daily sales
for one of the company's stores. The company provided the ML specialist with sales data for this
store from the past 10 years. The historical dataset includes the total amount of sales on each day for
the store. Approximately 10% of the days in the historical dataset are missing sales data.
The ML specialist builds a forecasting model based on the historical dataset. The specialist discovers
that the model does not meet the performance standards that the company requires.
Which action will MOST likely improve the performance for the forecasting model?
D vs C? Filling missing data with linear interpolation (D) ensures the model sees consistent input, which is usually better than just switching to weekly frequency (C) that might hide important daily patterns.
C changing to weekly might reduce noise and improve model stability.
D Filling those missing values with linear interpolation seems like the best first step since 10% missing data is pretty significant. If the model is trained on incomplete sequences or gaps, its accuracy will definitely suffer. Aggregating or smoothing might help later, but fixing the missing data issue directly addresses a clear problem in the dataset that could improve the model's learning. Changing to weekly forecasts might lose important daily patterns too.
D Using linear interpolation to fill missing sales data could help since 10% missing is quite a lot and might confuse the model. Fixing gaps often improves forecast accuracy more than just smoothing or aggregation.
D seems solid—filling missing days keeps daily granularity without adding noise.
It’s C. Switching to weekly sales reduces noise and the impact of missing daily data, making patterns clearer and improving forecast accuracy.
Does the question say why the model’s not performing well? Missing data is just 10%.