 |
 |
| Unit 5 ~ Quantitative Techniques for Business | |
| Lesson 8 - 28th March 2000 | |
| Time Series Analysis | |
Click here to see the slides from the lesson.
Download the notes
for this lesson in pdf format (1086Kb).
Download the
Toys are Back in Town Exercise in pdf format (149Kb).
Download the
To Infinity and Beyond Exercise in pdf format (175Kb).
- Moving averages,
- centred trend,
- seasonal variations
- Seasonally adjusted data using additive & multiplicative model
A time series is the name given to data collected over a period of time.
With the data in this form, we can plot a graph of time against takings. Time is usually on the horizontal axis.
Uses of Time Series
The use of collecting time-series data is to predict future events.
Short Term Variation
- Seasonal Fluctuations
- data can vary depending on the season (more people go to cinema at weekends)
- Cyclical Fluctuations
- data can vary due to semi-regular influences (business activity is affected by boom and slump)
- Residual Fluctuations
- data can also be affected by other events (flooding, strikes, fire, etc)
Moving Averages
- The Long Term Tendency is called the Trend.
- From a graph it can be impossible to quantify a trend because of seasonal variations
- In order to remove seasonal variations and quantify the trend, we use moving averages.
Predictions
- Using Time Series to make predictions requires caution, there are two assumptions:
- continuance of trend pattern
- assuming future behaviour will follow the same pattern as past behaviour
- no residual or cyclical fluctuations expected
- assuming no other influences, no economic or weather problems...
Copyright: All Content copyright of ©2000