Causal network generation system and data structure for causal relationship
A causal relationship is represented with a data structure including target identification information for identifying a target, index identification information of each of indexes to be used for quantitatively describing an event that occurs to the target, and causal relationship information for each pair of two indexes selected from the indexes. The causal relationship information includes direction information representing which one of the two indexes is a cause index and which one is an effect index, strength information representing a causal strength between the cause index and the effect index, time information representing a delay time for propagation of an influence of the cause index to the effect index, and correlation information representing a direction of change in the effect index with respect to an increase or decrease in the cause index.
1. A causal network generation system to generate a causal network representing a causal relationship between a plurality of events that occur to a target, the causal network generation system comprising a processor configured with a program to perform operations comprising:
operation as a data acquisition unit that acquires time series data of each of a plurality of indexes used for quantitatively describing the plurality of events;
operation as a causal relationship evaluation unit that evaluates, for each pair of two different indexes selected from the plurality of indexes, a causal relationship between the two indexes; and
operation an output unit that outputs data describing a causal relationship for each pair of the two indexes evaluated by the causal relationship evaluation unit, wherein the processor is configured to perform operations such that:
operation as the causal relationship evaluation unit comprises operation as the causal relationship evaluation unit that assumes one of the two indexes as a cause index and the other as an effect index, and calculating a causal strength between a change in the cause index and a change in the effect index after a time s while varying a value of the time s to evaluate a causal strength between the cause index and the effect index and a delay time for propagation of an influence of the cause index to the effect index,
operation as the causal relationship evaluation unit comprises operation as the causal relationship evaluation unit that calculates, as the causal strength, a transfer entropy TE XY (s) defined by an equation:
T
E
X
Y
(
s
)
=
[
log
2
P
(
y
(
t
+
s
)
,
y
(
t
)
,
x
(
t
)
)
·
P
(
y
(
t
)
)
P
(
y
(
t
)
,
x
(
t
)
)
·
P
(
y
(
t
+
s
)
,
y
(
t
)
)
]
where
t: time,
s: delay time,
x(t): time series data of the cause index,
y(t): time series data of the effect index,
P( ): probability density function, and
[*]: time average of *.
2. The causal network generation system according to claim 1 , wherein
the processor is configured with the program to perform operations such that operation as the causal relationship evaluation unit comprises operation as the causal relationship evaluation unit that evaluates a delay time with which the causal strength between the cause index and the effect index becomes maximum, and
the causal relationship output from the output unit includes information of the delay time with which the causal strength between the cause index and the effect index becomes maximum and information of a maximum causal strength between the cause index and the effect index.
3. The causal network generation system according to claim 1 , wherein
the processor is configured with the program to perform operations such that operation as the causal relationship evaluation unit comprises operation as the causal relationship evaluation unit that evaluates the causal strength between the cause index and the effect index for n delay times from first to n-th (where n is an integer greater than or equal to 2), and
the causal relationship output from the output unit includes information of n delay times from first to n-th and information of n causal strengths corresponding to the respective first to n-th delay times.
4. The causal network generation system according to claim 1 , wherein
the processor is configured with the program to perform operations such that operation as the causal relationship evaluation unit comprises operation as the causal relationship evaluation unit that calculates a correlation coefficient between a value of the cause index and a value of the effect index after the delay time, and
the causal relationship output from the output unit includes information of the correlation coefficient.