Embedded Files
What is time series all about?
Using data to make predictions (forecasting).
The data is collected at regular intervals of time.
This could be hourly, daily, monthly, yearly etc.
Finding the best model that fits this data.
Improving this model by better understanding its features and limitations.
An Introduction to Time Series
An Introduction to Time Series
Example of a time series analysis
Example of a time series analysis
Investigating the area of sea ice in polar regions.
Sea ice time lapse video
Looking at the data we see the following pattern:
The Process
The Process
- General trend
What do we notice?
Is there a general (long term) trend?
This could give us a general idea of what values we could expect in the future.
2. Short term patterns
2. Short term patterns
Are there any short term patterns?
This would improve our predictions we calculated using the long term trend.
3. Constructing a model
3. Constructing a model
Superimposing these two patterns provides us with a general model that we would use to make our predictions.
4. Forecasting
4. Forecasting
Analysis Overview
Analysis Overview
The components of a time series:
The components of a time series:
This standard involves using the statistical enquiry cycle to generate a model for a time series and then using it to make predictions.
You will use your understanding of the time series model to describe the accuracy (uncertainty) of your predictions.
Google Sites
Report abuse