 |
 |
 |
|
 |
|
 |
|
|
 |
 |
 |
Glossary
- Allometric function:
- A regression function that describes the growth rate of a part with respect to the growth of the entire organism (see Allometry)
- Allometry:
- Allometry is a concept developed in biology. “Allometry: the relative growth of a part in relation to an entire organism or to a standard” (Merriam-Webster)
- Auto-association:
- A procedure to measure the time dependancy of a phenomenon from a time-series compared to itself at different time lags. This technique is adapted for data measured at nominal level. (see Lag, Auto-correlation coefficient)
- Auto-correlation (temporal):
- A procedure to measure the time dependancy of a phenomenon from a time-series compared to itself at different time lags. (see Lag, Auto-correlation coefficient)
- Auto-correlation coefficient:
- A coefficient that expresses the correlation value between a time-series and itself at different time lags. Their scale of measurement must be at cardinal or ordinal level (see Lag, Cross-correlation)
- Change index (global):
- A change index is an indicator derived from multitemporal measurements. It expresses the amount of change within a period of time. It can describe the change behavior of a set of features (global) or of individual features. It can result from a difference, a ratio, ...
- Change vector analysis (CVA):
- The characteristics of a change in property value between two dates (moments) for 2 phenomena (variables) can be described as a vector expressing the strength (magnitude) of change as well as the direction of change with respect to the two variables
- Correlogramme:
- A graphical representation of a succession of correlation values varying throughout time or space (see Lag, Auto/Cross-correlation)
- Cross-association:
- A procedure to measure the correlation value between two time-series at different time lags. Their scale of measurement must be at nominal level (see Lag, Cross-correlation)
- Cross-correlation:
- A procedure to measure the correlation value between two time-series at different time lags. Their scale of measurement must be at cardinal or ordinal level (see Lag, Cross-association)
- Lag:
- In time series analysis, variables can be compared synchronously or with a defined time lag. As time series are generally made of regularly distributed intervals of time, this asynchronous comparison corresponds to one or several lag steps
- Linear regression function:
- A regression function that relates a dependant variable Y with one or several independant variables Xi in a linear manner. A first degree polynomial function is a linear function (see Polynomial regression function)
- Markov chain (analysis):
- A technique to estimate the probability of occurrence from any original state to any final state after a specific sequence of n time steps. It makes use of transition matrices
- Periodicity:
- A succession of properties that occurs regularly throughout time. For example daily temperatures or seasonal unemployement
- Principal component analysis (PCA):
- A procedure that transforms an original set of variables into a set of Principal components. This transformation removes the original correlation between variables (information redundancy) and structure the overall variability into ordered components (the first component carrying more variability than the second, and so on)
- Runs test:
- A runs test aims to compare an observed time series with a random sequence of states
- Similarity index:
- An index expressing the degree of similarity or association between two time series or one by itself at different lag positions (see Auto-association)
- Thematic properties:
- Values attached to observations expressing their property for each considered phenomenon (variable)
- Time series:
- A sequence of measurements ordered according to Time (moments of time). It describes the change of properties of a single observation throughout time
- Transition matrix:
- A general term to identify any matrix that expresses a change of properties (states) between two moments
- Transition probability matrix:
- A probability matrix that expresses a transition from one state to another
- Transition proportion matrix:
- A matrix that expresses the tendency of one state to follow another
- Trend surface (analysis):
- A regression function modelling the property values (Z) based on their location (X,Y) in space: Z = f(X,Y)
 |
|
|
|
|
|
|
|
