Which method is described as better than the other two time-series methods?

• A. Naive forecast
• B. Cumulative mean
• C. Moving average
• D. Weighted mean

Answer: (C)

In time-series forecasting, smoothing approaches that average recent observations tend to outperform very simple baselines because they reduce random noise while staying responsive to recent changes.

A moving average forecast averages a fixed number of most recent observations. This dampens random fluctuations, so the forecast reflects recent behavior without being tied to a single past value. That balance often yields more accurate short-term predictions.

The naive forecast, which simply repeats the last observed value, can be pulled off course by sudden shifts or noise and fails to anticipate changes in direction. The cumulative mean takes every past observation into account equally, which makes it slow to adapt when the series shifts; it lags behind new patterns and trends.

Because of these differences, the moving average typically provides a better forecast of the near term than both the naive approach and the cumulative mean, especially when the data have short-term fluctuations but some underlying continuity. The effectiveness depends on choosing an appropriate window length: too small a window behaves like the naive forecast, while too large a window can over-smooth and miss recent changes.