Operations Management Concepts: Forecasting, Quality, and Inventory
Q1. Time Series Forecasting Methods Comparison
Time Series Components
- Trend (T)
- Seasonality (S)
- Cyclical (C)
- Random (R)
Forecasting Methods
Naïve Method
Formula: Ft+1 = At
- Assumes next demand equals last demand.
- Used as a benchmark.
Simple Moving Average (n)
Formula: Ft+1 = (At + At-1 + … + At-n+1) / n
- Smooths random variation.
- A higher 'n' increases lag.
Weighted Moving Average
Formula: Ft+1 = w1At + w2At-1 + … + wnAt-n+1
- Sum of weights (Σw) equals 1.
- Recent data is given a higher weight.
Exponential Smoothing
Formula: Ft+1 = αAt + (1−α)Ft
- Where 0 < α < 1.
- A higher α increases responsiveness.
- Most widely used method.
Comparison Summary
The Naïve method is the simplest. The Moving Average (MA) smooths noise. The Weighted Moving Average (WMA) is more...