Topic 15 Flashcards
How can forecasting be defined?
Forecasting is a technique that is used to predict, with some measure of probability, what is likely to happen to a time series in the future, on a basis of past experience or past relationships with other time series.
What does the accuracy of forecasting depend on?
The accuracy of forecasting depends on the accuracy of past data and the selected forecasting technique.
What are the principles of forecasting?
The principles of forecasting are: the future period should be similar to the past period, enough data should be used, and test forecasts using different methods should be compared with actual data.
What is direct data?
Direct data is the time series of information of interest.
What is indirect data?
Indirect data is information that can explain anomalies in the direct data.
What is time series forecasting?
Time series forecasting approach seeks to forecast future values using only direct data.
What is econometric forecasting?
Econometric forecasting seeks to forecast future value of a time series using both direct and indirect data.
What are the typical data patterns in time series data?
Typical data patterns in time series data are trend, cyclical, seasonal, and random.
What is a trend in time series data?
Trend refers to an upward or downward pattern over a longer time period in time series data.
What is a cyclical pattern in time series data?
Cyclical patterns in time series data are influenced by economic factors.
What is a seasonal pattern in time series data?
Seasonal patterns in time series data are influenced by factors such as weather, holidays, etc. They typically follow an annual cycle.
What is a random pattern in time series data?
Random patterns in time series data are due to unpredictable events.
What is the Moving Average method in time series forecasting?
The Moving Average method is a time series forecasting technique where the forecast for the next period is the average of the previous ‘n’ periods. It is a simple technique that helps smooth out short-term fluctuations to highlight longer-term trends or cycles.
What is Exponential Smoothing in time series forecasting?
Exponential Smoothing is a time series forecasting method that involves calculating the forecast for the next period combining the previous forecast and the previous actual value, with the weights declining exponentially. It gives more weight to recent observations and less weight to older ones, hence the term ‘exponential’. It is particularly effective for data with no trend or seasonal pattern.
