Forecasting
Predicting the future based on past data. Essential for inventory, sales, and budgeting.
Components of a Time Series
- Trend (T): Long-term upward or downward movement.
- Seasonality (S): Regular, repeating patterns (e.g., holiday sales).
- Cyclical (C): Long-term wave-like patterns (e.g., business cycles).
- Random (R): Unpredictable noise.
Smoothing Methods
Used to remove random noise to see the trend.
Moving Average (MA)
Average of the last $k$ periods.
$$ MA_t = \frac{\sum \text{last } k \text{ values}}{k} $$
Exponential Smoothing
Gives more weight to recent data.
$$ F_{t+1} = \alpha Y_t + (1-\alpha)F_t $$
$\alpha$ = Smoothing constant (0 to 1)
$\alpha$ = Smoothing constant (0 to 1)
Trend Projection
Using linear regression where $X$ is time ($t$).
$$ \hat{Y} = b_0 + b_1t $$
Test Yourself
Q1: Which component represents regular, repeating patterns within a year?