Book - Chapter 8 Analytical Theory Time Series Flashcards
What does timeseries mean
Ordered sequence of equally spaced values over time
What does timeseries analysis mean
Accounts For the internal structure of observations taken over time
What are the goals of timeseries analysis
To identify the internal structure of the time series. To forecast future events
What is the box Jenkins method
A RMA. Autoregressive moving averages
What does the box Jenkins method do
Models historical behaviour to forecast the future
What would you use box Jenkins Method /ARMA to do
Forecast. For example, next months sales, tomorrow’s stock price, hourly power demand
What three components do you need to model a timeseries
The trend term. The seasonal term. The random component
What does box Jenkins methodology assume
The random component is a stationary sequence
What makes a stationary sequence
Constant mean. Constant variance. Autocorrelation does not change over time
To obtain a stationary sequence what must the data be
Detrended and seasonally adjusted
How would you remove a simple linear trend
By subtracting a least squares fit straight line
ARMA model is also known as
The P, Q model
ARIMA as what to the ARMA model
Adds a difference in time
In Arima what does it mean
Autoaggressive intra dated moving average. It includes the detrending as part of the model
what is ARIMA also known as
The p d q model.
The Arima model, Multiple models need to be built in compared using
ACF and PACF
What is ACF
Autocorrelation function. Correlation of the values of the timeseries with itself. It helps to determine the order, Q, of a MA model
What is PACF
Partial autocorrelation function. And autocorrelation calculated after removing the linear dependence of the previous terms. Helps to determine the order, P, of a AR model
What is the quantity H in ACF referred to as
Lag
What is the white noise process
The Eire term and is used to represent random, independent fluctuations that are part of the timeseries
What is the PACF
After linear regression is used to remove the effect of the variables the PaCF is the correlation of what remains