Föreläsningar del 1 Flashcards
What are quantitative forecasting methods based on?
Historical data concerning one or more time series.
What is a time series?
A set of observations measured at successive points or periods in time.
What is the difference between time series methods and causal methods?
Time series methods use past values of the same series, while causal methods use related time series.
What are the three main time series methods?
Smoothing, trend projection, and trend projection adjusted for seasonality.
What does the trend component of a time series account for?
The gradual shift of the series over a long period.
What does the seasonal component represent?
Regular patterns of variability within certain time periods (e.g., yearly cycles).
What is the irregular component of a time series?
Short-term, unpredictable variations that cannot be forecasted in advance.
What is Mean Squared Error (MSE)?
The average of squared forecast errors; the method that minimizes MSE is preferred.
What is Mean Absolute Deviation (MAD)?
The mean of absolute forecast errors; it is less sensitive to large errors than MSE.
When are smoothing methods useful?
When the time series is stable without significant trends, seasonal, or cyclical effects.
What are the four common smoothing methods?
Moving averages, centered moving averages, weighted moving averages, and exponential smoothing.
How does the moving average method work?
It averages the most recent nnn data points to forecast the next period.
What is the centered moving average method used for?
Computing season indexes by averaging nnn periods and assigning it to the midpoint.
How does the weighted moving average method differ from the simple moving average?
It assigns more weight to recent observations, with weights summing to 1.
What is the principle of exponential smoothing?
The forecast for the next period is the current forecast plus a proportion α\alphaα of the forecast error.
