JP2007265229A - Time series analysis program, time series analysis system and terminal unit used for the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a program and system for making a time series analysis of a model of an influence diagram, and a terminal unit used for the same. <P>SOLUTION: The time series analysis program 121, provided in the terminal unit 100 of the time series analysis system 1, acquires node setting information for every repetition time corresponding to time series, and stores node names and repetition times in relation to each other in a node information storage section 131. A low-order node and a node of the last repetition time can be set in a mathematical expression of node setting information. In simulation processing by the time series analysis program 121, node values for every repetition time are computed using node setting information and stored in a node value information storage section 151. Using the node values stored in the node value information storage section 151, an expected value, a random variable, standard deviation, a confidence interval upper limit value and a confidence interval lower limit value are computed and displayed as a simulation result of the time series analysis. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a model in which a plurality of nodes are connected by a relation line to indicate a mutual relationship, and more particularly, to a technique for time-series analysis of a model created with an influence diagram.

An influence diagram is used as a method for diagramming a business model and quantifying the causal relationship of influence on an evaluation target. At that time, it is possible to perform quantitative analysis of the model by setting mathematical formulas including random variables in association with each node of the model created on the influence diagram and obtaining the evaluation value of the entire model by simulation processing it can.
An analysis method using an influence diagram is generally used for evaluating a business model of a company and analyzing a company value.

Technologies related to influence diagrams include the following.
Non-Patent Document 1 describes the definition of nodes of an influence diagram and a method of calculating an actual evaluation value as a technique related to an evaluation method of an influence diagram.
Non-Patent Document 2 proposes an evaluation / analysis method using Monte Carlo simulation for making a decision using an influence diagram.
Non-Patent Document 3 describes a method of sensitivity analysis using an influence diagram.
ROSS D.SHACHTER, "EVALUATING INFLUENCE DIAGRAMS", "Operation Research", November-December 1986, Vol 34, No.6 Concha Bielza, Peter Muller, David Rios Insua, "Decision Analysis by Augmented Probability Simulation", "Management Science", July 1999, Vol. 45, No. 7 Thomas Dyhre Nielsen and Finn V. Jensen, “Sensitivity Analysis in Influence Diagram”, “IEEE Transactions on Systems, Man, and Cybernetics.Part A: Systems and Humans.”, March 2003, Vol. 33, No. 1.

  Since analysis of models using influence diagrams does not relate information from multiple influence diagrams, it is possible to obtain only an evaluation index at a certain point in time, and it is assumed that time series models will be drawn and analyzed. Not. Therefore, when analyzing a time-series model using an influence diagram, it is necessary to analyze a model for each unit period at a certain point by manually drawing them one by one and combining them manually. there were.

  Therefore, in view of the above problems, the present invention has an object to provide means for efficiently performing analysis based on information of a plurality of influence diagrams such as time series analysis in a system that performs model analysis with influence diagrams. To do.

In order to solve the above problems, one of the solving means according to the present invention is a time series analysis program for performing time series analysis of a model in which a plurality of nodes are connected by a relation line to show a mutual relationship, and a computer The model information acquisition means for acquiring the node information of the model input via the input means and the information of the relation line indicating the relationship between the nodes, the information of each node input via the input means Node setting information is acquired, node setting information acquiring means for associating the node setting information with a node name and a repetition time in a time series, and information on the upper limit of the number of repetitions input via the input means, and the acquisition Using the node information and the related line information, the value of each node for each iteration is calculated until the iteration count reaches the iteration count upper limit value. And simulation means that the value of each node in each repetition times that the calculated, said the repeated times order, characterized in that to the simulation result displaying means functions as, for time-series display as a simulation result.
Other means will be described in an embodiment described later.

  ADVANTAGE OF THE INVENTION According to this invention, in the system which performs a model analysis with an influence diagram, the analysis based on the information of several influence diagrams, such as a time series analysis, can be performed efficiently.

The best mode for carrying out the present invention (hereinafter referred to as “embodiment”) will be described in detail.
In the description of this embodiment, “time series analysis system” is abbreviated as “system” as appropriate, and “time series analysis program” is abbreviated as “program” as appropriate.

  As described above, the influence diagram is a technique for graphically modeling a business model and quantitatively analyzing the causal relationship of influence on an evaluation target, and is used for evaluating a business model of a company, analyzing a corporate value, and the like.

The configuration of the influence diagram will be described.
The influence diagram is composed of four elements: an evaluation value node, a random variable node, a decision node, and a relation line. The evaluation value node is a node indicating the final evaluation value of the influence diagram, and corresponds to the root node of the influence diagram. The value of this evaluation value node becomes the final evaluation value of the model created by the influence diagram. The random variable node is a node that defines a probability distribution using lower nodes. The decision node is a node to which a value to be determined by the decision maker is input in the influence diagram. The relation line is an arrow line expressing a relation between nodes. The arrow line of the relation line shows the dependency relationship of each node, and represents the relationship that the node value of the original node (lower node) of the arrow affects the node (upper node) of the arrow destination. Yes.
The influence diagram is expressed as an effective graph in which nodes are connected with relational lines.

FIG. 1 is a diagram showing an example of an influence diagram model.
In FIG. 1, a “profit” node 201 that is an evaluation value node of this model, a “cost” node 202 and a “sales” node 203 that are random variable nodes, a “man-hour” node 204 that is a decision-making node, and a “unit price”. Node 205 and associated lines 206 connecting each node are shown.
In the present specification, the model of the influence diagram in FIG. 1 will be described as an example.
In addition, although the example of this specification demonstrates as one evaluation value node, it is not restricted to this, A plurality of evaluation value nodes may be set.

<System configuration>
FIG. 2 is a block diagram illustrating an example of the configuration of the system 1.

The time series analysis system 1 is realized by the terminal device 100, the input device 180, and the output device 190.
The terminal device 100 is a general computer, and includes a user interface unit 110 for interactively creating an influence diagram model by a user operation, a storage unit 120 that is a nonvolatile storage device such as a hard disk device, and various operations. It is mainly composed of a processing unit 160 that executes processing, a RAM 170 that is a temporary storage area used for arithmetic processing, and the like.

  The user interface unit 110 controls an input device 180 such as a keyboard and a mouse connected to the terminal device 100 and an output device 190 such as a display, and provides a user interface for interactively creating an influence diagram model. . Specifically, the user interface unit 110 displays the processing result obtained by the implementation of the system 1 on the output device 190 and inputs from the input device 180 according to the input item from the displayed processing result screen. The function of receiving the received data and transmitting it to the storage unit 120 or the processing unit 160 is provided.

The storage unit 120 includes a time series analysis program 121, a node information storage unit 131 that is data handled by the system 1, a related line information storage unit 141, and a node value information storage unit 151.
The time series analysis program 121 is loaded into the RAM 170 as necessary and executed by the processing unit 160.

  The processing unit 160 is a central processing unit of the terminal device 100, and executes a time series analysis program 121 loaded in the RAM 170, thereby executing a simulation unit (simulation unit) 161 and a model information acquisition unit (model information acquisition unit) 162. The functions of the node setting information acquisition unit (node setting information acquisition unit) 163 and the simulation result display unit (simulation result display unit) 164 are realized.

  The simulation unit 161 has a function of obtaining an evaluation value of the model of the influence diagram expressed by information held in the node information storage unit 131 and the related line information storage unit 141. Note that the function of the simulation unit 161 may be realized by using spreadsheet software.

  The model information acquisition unit 162 has a function of displaying an influence diagram model and acquiring information of the influence diagram model drawn via the input device 180.

  The node setting information acquisition unit 163 has a function of providing the user interface unit 110 with a user interface for displaying and editing each node setting information of the entire time series of each node.

The simulation result display unit 164 uses a value of each node as a simulation result stored in the node value information storage unit 151 to display a user interface for displaying a probability distribution of each node and a time series transition of evaluation values. It has functions such as providing it to the interface unit 110.
Details of each of these functions will be described later with reference to FIGS.

(Node information)
FIG. 3 is a diagram illustrating a configuration example of the node information storage unit 131 in which node setting information is stored.
The node information storage unit 131 includes a node name 310 for uniquely identifying a node, and node setting information for repeatedly storing information set in the node every 311 (1 to n). The information set in the node setting information is stored, for example, for each node as n sequential numbers starting from 1. n corresponds to the upper limit of the number of repetitions.
For example, when the time series analysis is performed every month for one year, the upper limit “n” of the number of repetitions is “12”, and the monthly repetition number of 1, 2, 3,. Become.

  In the node setting information, in addition to the value, mathematical expressions such as four arithmetic operations and functions can be set. At this time, the value of the node name 310 of the lower node of the node storing the mathematical expression is used. Thus, the value of the lower node of the repetition times 311 (1 to n) can be used.

  Specifically, referring to FIG. 3, for example, for a record whose node name 310 is “profit” (hereinafter referred to as “profit node” as appropriate), the repeat count 311 is “1” (hereinafter “repeat count”). In the node setting information (denoted as “1” as appropriate), information on “sales-cost” is stored. This indicates that the value of “cost node” is subtracted from the value of “sales node”, which is a subordinate node of “profit node”, in the repetition “1”.

  In addition, the value of each node of the repetition times immediately before the repetition times 311 (1 to n) can be used for the mathematical expression of the repetition times “2” or more (second and subsequent times). In that case, the time series analysis program 121 recognizes it by adding “$” to the head of the node name specified in the mathematical expression.

  More specifically, for example, node setting information of “2” repeatedly in “profit node” stores information on “sales−cost + $ profit”. According to this, the value of the “cost node” is subtracted from the value of the “sales node” of the repeated times “2”, and the value of the “profit node” of the repeated times “1” is added to the value. Indicates.

(Related line information)
FIG. 4 is a diagram illustrating a configuration example of the related line information storage unit 141.
In the related line information storage unit 141, a related line name 410 for uniquely identifying a related line, a connection source node name 411 in which a connection source node (lower node) is set, and a connection destination node (upper node) are set. And a connection destination node name 412.

(Node value information)
FIG. 5 is a diagram illustrating a configuration example of the node value information storage unit 151 in which the value of the node is stored.
The node value information storage unit 151 includes a node name 510, a repetition count 511, and sample values 512 (1 to m).
Each record in the node value information storage unit 151 is uniquely identified by using the node name 510 and the repetition times 511 as keys, and the sample value 512 holds the value of the node calculated by the simulation processing for each number of simulations. The number of simulations corresponds to the number of samples for the system 1 to calculate a standard deviation displayed as a result of executing the simulation process.

<Process overview>
The procedure of the time series analysis of the model using the influence diagram in this system will be described based on the flow shown in FIG. 6 and with reference to FIGS.

  First, the simulation unit 161 acquires a time-series repetition count upper limit value input by the user via the input device 180 (S600). The value acquired here is used as the number “n” of data of the node setting information in the node information storage unit 131, that is, the number of node setting information that can be set in the node.

Next, the user draws a model via the input device 180 (node generation, nodes are connected by related lines), and the model information acquisition unit 162 displays the screen on the output device 190 using the user interface unit 110, Node and related line information is acquired (S601) and stored in the node information storage unit 131 and the related line information storage unit 141.
FIG. 9 is a diagram illustrating a screen display example displayed on the output device 190 by the model information acquisition unit 162. In the related line information storage unit 141 in FIG. 4, data based on the example in FIG. 9 is shown.

  The node setting information acquisition unit 163 acquires node setting information for each repetition 311 (1 to n) of the node information storage unit 131 for each node of the model drawn in step S601 (S602). Specifically, the node setting information for each repetition 311 (1 to n) of each node input by the user via the input device 180 is acquired and stored in the node setting information in the node information storage unit 131.

An example of an input screen for node setting information displayed on the output device 190 is shown in FIG.
In FIG. 10, the left side of the screen is a “diagram display screen” 1010 in which an influence diagram model is displayed. The right side of the screen is a “node setting information input screen” 1020, which is a screen for inputting node setting information of a selected node among the models displayed on the left side of the screen. The numerical values displayed in the model node of the “diagram display screen” 1010 in FIG. 10 will be described later.

In FIG. 10, since the “profit” node is selected on the “diagram display screen” 1010, “profit” is displayed in the area 1021 of the “node setting information input screen” 1020. A region 1022 indicates the number of repetitions. The radio button 1023 can select whether the information input in the frame of the area 1024 is a calculation formula or a conditional formula. An area 1024 is a field for inputting node setting information. If the node name is included in the mathematical expression input in the frame of the area 1024, the user selects a desired node name from the pull-down menu 1025 via the input device 180 and presses the “Add Distinguished Name” button 1026. As a result, the node name is displayed within the frame of the area 1024. Note that the user may input the node name directly in the frame of the area 1024 using the input device 180 such as a keyboard.
When “conditional expression” is selected with the radio button 1023, a conditional expression such as an IF statement can be input in the frame of the area 1024.

  It is possible to input node setting information for each repetition by switching the tab 1011 from “diagram” to “list”.

FIG. 11 is an example of a screen displayed on the output device 190 when the tab 1011 in FIG. 10 is switched to “list”.
In FIG. 11, the left side of the screen is a “list screen” 1110 and the values of each node for each repetition are displayed in a list format. The right side of the screen is a “node setting information input screen” 1020 as in FIG. ing.

Here, in the “list screen” 1110, the repetition number “2” of the “profit” node is selected as shown in the area 1111, so that in the “node setting information input screen” 1020, the area As indicated by 1122, node setting information of the repetition number “2” of the “profit” node can be input within the frame of the area 1124.
In FIG. 11, the numerical value in the frame displayed in the table format of the “list screen” 1110 on the left side of the screen is the node value for each repetition of each node calculated using the node setting information on the right side of the screen. It is displayed.

  In addition, in the node setting information stored in the node value information storage unit 151, the node setting information for which there is no input, the node setting information of the repetition count “1” of the node is used, or there is an input at the node. The specification may use the node setting information repeatedly.

The node setting information acquisition unit 163 may display a tabular screen (portion of “list screen” 1110 in FIG. 11) for inputting the node setting information in a separate window as shown in FIG. Similarly, only the portion of the screen “diagram display screen” 1010 in FIG. 10 may be displayed in a separate window as shown in FIG.
Although not shown, node setting information may be displayed, input, and edited instead of the node value displayed in the item of “list screen” 1110 in FIG. Thereby, the user can input node setting information including mathematical expressions in a list format.

  When correction is necessary for the upper limit “n” of the number of repetitions, the model, the node setting information, etc., the process returns to S600, S601, or S602 as appropriate (not shown).

Returning to FIG.
When the drawing of the model and the acquisition of each node setting information are completed, the simulation unit 161 acquires the number of simulations (“m” in FIG. 5) input by the user via the input device 180 (S603). As described above, the value acquired here is the number of times of sampling (the number of samples) during simulation execution.

  Thereafter, the simulation unit 161 executes a simulation process (S604). The detailed procedure of the simulation process will be described later with reference to FIG. Note that the values of all the nodes calculated by executing the simulation process are stored in the node value information storage unit 151 in FIG.

Subsequently, the simulation result display unit 164 acquires the node specified by the user via the input device 180 and the information on the repetition times, and calculates the probability distribution, the expected value, the standard deviation, etc. for the node and the repetition times. Then, it is displayed on the output device 190 via the user interface unit 110 (S605). That is, the simulation result display unit 164 determines the probability distribution and expected value of the node value of the repeated times at the node from the set of node values stored in the sample values 512 (1 to m) of the node value information storage unit 151. And the standard deviation is calculated and displayed.
The simulation result display unit 164 displays the calculated values in a histogram format or the like via the user interface unit 110.
FIG. 13 is a diagram illustrating a screen display example in the histogram format of the probability distribution.
This screen shows the probability distribution / expected value / standard deviation information calculated in step S605. The user selects a node and repetition times on the “diagram display screen” 1010 in FIG. By selecting “Display”-“Histogram” in the bar, the screen of FIG. 13 is displayed. The same applies to the case where a node and repetition times are selected on the “list screen” 1110 in FIG. 11 and “display”-“histogram” is selected from the menu bar.
In FIG. 13, the node name designated by the user is displayed in the area 1301, and the repeated times designated by the user are displayed in the area 1302. Further, the probability distribution for the node and the repetition times is displayed as a histogram, and the expected value and the standard deviation are displayed as numerical values.
As a result, the accuracy of the evaluation value for the repeated times of the node can be visually grasped by the inclination of the histogram.
Note that the expected value and the standard deviation may be displayed on the output device 190 in a table format or the like by calculating the probability distribution, the expected value, and the standard deviation for each iteration at the node.

  The simulation result display unit 164 performs time-series transition display processing by displaying the expected value and standard deviation of the evaluation value node on the output device 190 via the user interface unit 110 as time-series transitions (S606). ). A detailed description of the time series transition display process will be given later in FIG.

  The user determines whether to review the model again based on the simulation processing result displayed on the screen, and if the review is necessary, returns to S600, S601 or S602 to review the model (not shown). ).

<Simulation process>
The influence diagram simulation process according to the present embodiment will be described with reference to FIGS. This process corresponds to the process of step S604 in FIG.

  First, the simulation unit 161 uses the related line information storage unit 141 to topologically sort each node of the simulation target model based on the evaluation value node, and calculates the calculation order of the node values (S700).

In the simulation process, the simulation unit 161 performs a triple loop process.
In the outermost loop, a loop of input time series repetitions (up to the repetition number upper limit “n”) is performed (S701). By this loop, each node value corresponding to each repetition (1 to n) is calculated.
In the second loop, a loop for the number of simulations (the number of samples “m”) is performed (S702). By this loop, each node value corresponding to each simulation count (sample number “m”) is calculated.
In the innermost loop, a loop is performed for all the nodes in the order sorted by the topological sort until the evaluation value nodes are obtained (S703). By this loop, node value calculation processing for all nodes is performed.
The node value calculation process is performed according to the following procedure.

  First, the simulation unit 161 acquires node setting information corresponding to the current iteration set for the node from the node information storage unit 131 (S704).

  The simulation unit 161 determines whether or not the current node number of the current iteration is present in the node setting information formula acquired in step S704, and if it exists (S705 → Y), the current iteration is determined. And the node value of the said lower node corresponding to the present simulation frequency is acquired from the sample value 512 (1-m) of the node value information storage part 151 (S706). On the other hand, if it does not exist (S705 → N), the process proceeds to step S707.

  The simulation unit 161 determines whether or not the node name of the previous iteration is present in the node setting information obtained in step S704, and if it exists (S707 → Y), The node value of the corresponding node in the previous (previous) iteration is acquired from the sample values 512 (1 to m) in the node value information storage unit 151 (S708). On the other hand, if it does not exist (S707 → N), the process proceeds to step S709.

Thereafter, the simulation unit 161 calculates a node value using the node value setting information acquired in step S704 (S709), and the calculated node value is stored in the node value information storage unit 151 in the current node. Is stored in a sample value 512 corresponding to the current iteration number (S710).
By performing the above simulation processing, the node value information storing unit 151 stores the node value information using the node name and the repetition times as the keys in the sample values 512 (1 to m).
Here, the simulation unit 161 stores each node value stored in the sample value 512 (1 to m) of the last iteration number that matches the upper limit value of the iteration number stored in the node value information storage unit 151. The expected value calculated from the set may be displayed on the screen of the output device 190. The numerical values displayed in the model of the “diagram display screen” 1010 in FIG. 10 correspond to this.
Furthermore, each value of the “list screen” 1110 in FIG. 11 and FIG. 12 are sample values 512 (1 to m) for each repetition of each node value stored in the node value information storage unit 151 in FIG. The expected value calculated from the set of node values stored in is displayed.

<Time-series transition display processing>
The display process of the time series transition of evaluation values in the present embodiment will be described with reference to FIGS. This process corresponds to the process of step S606 in FIG.

  First, the simulation result display unit 164 acquires the percentile of the confidence interval used for calculation of the confidence interval upper limit value and the confidence interval lower limit value input by the user via the input device 180 (S800).

  Next, the simulation result display unit 164 uses the node value information storage unit 151 until each node reaches the repetition count upper limit “n” (S801). 1 to m) is acquired, and the probability distribution, expected value, standard deviation, confidence interval upper limit value, and confidence interval lower limit value of the evaluation value node are calculated (S802).

  Thereafter, the simulation result display unit 164 sends the respective values such as the probability distribution, expected value, standard deviation, confidence interval upper limit value, confidence interval lower limit value, etc. for each iteration calculated in step S802 via the user interface unit 110. It is displayed on the output device 190 as a time series transition graph in order of repetition (S803).

A screen display example of a time-series transition graph of evaluation value nodes displayed on the output device 190 is shown in FIG. Here, the confidence interval represented by the confidence interval upper limit value and the confidence interval lower limit value is shown in a graph together with the expected value and standard deviation for each iteration.
When the user changes the confidence interval upper limit value or the confidence interval lower limit value based on the time-series transition graph displayed on the screen of the output device 190, the “confidence interval upper limit” in the area 1401 in FIG. The value of “confidence interval lower limit” is changed, and the update button 1403 is pressed. Accordingly, the “time-series transition display process” returns to step S800 in FIG. 8 to continue the process. Here, the histogram described with reference to FIG. 13 may be displayed by double-clicking a certain number of times of information (for example, an arrow 1404) of the node displayed in FIG.
Further, for each node other than the evaluation value node, a time series transition display device (steps S801 to S803) may be performed to display a time series transition graph. Thereby, the user can grasp the time series transition of nodes other than the evaluation value node.

The simulation result display unit 164 may display a node setting information input screen, a node value information display screen, a probability distribution histogram format display screen, an influence diagram time series analysis graph, and the like on the same screen. . FIG. 15 is a diagram in which each screen is displayed on the same screen in the output device 190. According to this, the nodes of the influence diagram model, the node setting information, and the like can be corrected, and the result can be immediately confirmed in a tabular form of a list or a time series transition graph.
As a result, the results of the time series analysis based on the information of the influence diagram can be provided to the user in an easy-to-understand manner by displaying the results in a time series in a graph format or a table format.
In addition, the influence diagram model and various analysis results can be displayed at the same time, and the simulation process can be executed immediately, so that it is possible to support the user's quick decision making.

According to the embodiment of the present invention described above, the node setting information to be set in the node is stored in association with the unique information indicating the repetition times, so that a plurality of different mathematical expressions can be stored in the nodes of the influence diagram model. Node setting information such as can be set.
Then, using the node setting information defined in the model, the evaluation value of the model for each iteration is obtained.
The formula included in the node setting information used when obtaining the evaluation value of the model is stored as information associated with the value of the current iteration count, and the formula includes the value of the lower node and the iteration immediately before the iteration count. You can use each node value obtained in the first time.
Among the obtained evaluation values, the node value of the evaluation value node obtained in the last iteration is used as the final evaluation value in the model time series, and the set of obtained evaluation values is arranged in the order of the iteration. Display things as time series of models.
In addition, a function for displaying the node value for each repetition of each node in a table format and a function for inputting model node setting information in a list format are provided.

By providing the above-described function, the following effects can be obtained.
In the influence diagram model, it is possible to use not only the value of the lower node but also the value of the calculation result of the previous iteration of each node in the expression of each node. Time series analysis of the model of the diagram can be easily performed in a short time.

  In addition, by creating an influence diagram model using GUI (Graphical User Interface), it is configured to have user operability, and on the same screen, the influence diagram model, node setting information, node Value and graph display can be changed. As a result, the result of changing the setting information of each node can be grasped by the user in real time both numerically and graphically, and effective verification and analysis can be performed more quickly. This function is effective as an execution decision material for management decisions, strategic meetings, and research and development.

  Since the result of calculating the standard deviation and the confidence interval on the influence diagram can be displayed on one screen (graph), the user can easily grasp the analysis result visually. Then, by changing and updating the confidence interval upper limit value and the confidence interval lower limit value on the display screen, the result is immediately reflected in the graph, so that the user can grasp the simulation processing result in real time.

  According to the above, it is possible to reduce the number of preparation work man-hours for performing the time series analysis of the influence diagram, and to facilitate the operation of the artificial judgment work.

  In the present embodiment, the time series analysis program 121 has been described as being stored in the storage unit 120 together with the data. However, the present invention is not limited to this. It may be stored. Further, the program 121 may be provided as a CD-ROM (Compact Disk Read Only Memory) or a semiconductor chip in which the program 121 is written. As for the program installation, the program may be downloaded and installed via a network via a communication device.

  The time series analysis system 1 is realized by a computer such as a personal computer or a server. However, the time series analysis system 1 does not necessarily have to be a stand-alone, and may be realized by a client / server system.

  As mentioned above, although an example was shown about suitable embodiment of this invention, this invention is not limited to the said embodiment, In the range which does not deviate from the meaning of this invention, it can change suitably.

It is a figure which shows the example model of an influence diagram. It is a block diagram which shows an example of the function of this system. It is a figure which shows the structural example of a node information storage part. It is a figure which shows the structural example of a related line information storage part. It is a figure which shows the structural example of a node value information storage part. It is a flowchart which shows the process sequence of the time series analysis in this system. It is a flowchart which shows the process sequence of a simulation process. It is a flowchart which shows the process sequence of the display process of a time series transition. It is a figure which shows the example of a screen display of an influence diagram. It is a figure which shows the example of an input screen of node setting information. It is a figure which shows the example of an input screen of node setting information on a "list screen". It is a figure which shows the example of another window display screen of a "list screen." It is a figure which shows the example of a screen display by the histogram format of probability distribution. It is a figure which shows the example of a screen display of the time series transition graph of an evaluation value node. It is a figure which shows the example which displayed the setting screen of node setting information, the display screen of node value information, the histogram format display screen of probability distribution, the time series analysis graph of an influence diagram, etc. on the same screen.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Time series analysis system 100 Terminal device 110 User interface part 120 Storage part 121 Time series analysis program 130 Data part 131 Node information storage part 141 Related line information storage part 151 Node value information storage part 160 Processing part 161 Simulation part 162 Model information acquisition Unit 163 Node setting information acquisition unit 164 Simulation result display unit 180 Input device 190 Output device

Claims (9)

A time series analysis program for performing a time series analysis of a model in which a plurality of nodes are connected by a relation line to indicate a mutual relationship,
Model information acquisition means for acquiring node information of the model input via the input means and information of relation lines indicating the relationship between the nodes;
Node setting information acquisition means for acquiring node setting information of each node input via the input means, and associating the node setting information with a node name and a repetition time in a time series;
The information on the upper limit of the number of repetitions input via the input means is acquired, and the number of repetitions is repeated until the number of repetitions reaches the upper limit of the number of repetitions using the acquired information on the node and information on the relation line. Simulation means for calculating the value of each node for each;
A simulation result display means for displaying the value of each node for each calculated repetition time in a time series as the simulation result in the repetition time order,
A time series analysis program characterized by functioning as The simulation means includes
When the node setting information includes information indicating a lower node, the value of the node is calculated using the information of the value of the lower node among the calculated values of each node for each iteration. The time series analysis program according to claim 1, wherein: The simulation means includes
When the node setting information includes information indicating the node of the previous iteration, the information on the value of the node of the previous iteration is calculated among the calculated values of each node. The time series analysis program according to claim 1, wherein the value of the node is calculated using the time series analysis program. The simulation result display means includes
The time series analysis program according to claim 1, wherein the information on the value of each node of the number of repetitions that matches the upper limit of the number of repetitions is displayed. The node setting information acquisition unit includes:
The time-series analysis program according to claim 1, wherein a tabular screen for inputting the node setting information is displayed. The node setting information acquisition unit includes:
The time series analysis program according to claim 1, wherein the calculated value of each node for each repetition is displayed in a table format. The simulation result display means includes
Using the calculated value of each node for each iteration, at least one of an expected value, a confidence interval upper limit value, a confidence interval lower limit value, and a standard deviation of each node is calculated, and the calculation result is used as the simulation result. The time-series analysis program according to claim 1, wherein the time-series analysis program is displayed on the same screen. A time-series analysis system comprising: a terminal device that analyzes a model created by an influence diagram; an input device that inputs information to the terminal device; and an output device that outputs information calculated by the terminal device. And
The terminal device
A node information storage unit for storing node setting information to be set in a node constituting the model in association with a node name and a repetition time in a time series;
A related line information storage unit that stores information of related lines indicating the relationship between the nodes;
A node value information storage unit for storing the value of the node in association with the node name and the information of the repetition times;
A processing unit that performs time series analysis using each information,
The processor is
Information on nodes constituting the model input via the input device and information on the related lines are acquired, information on the nodes is stored in a node information storage unit, and information on the related lines is stored in related line information. Model information acquisition means stored in the section;
Node setting information acquisition means for acquiring the node setting information input via the input device and storing the node setting information in the node information storage unit;
Obtaining information on the upper limit of the number of repetitions input via the input device, using the information of the node information storage unit and the related line information storage unit, until the repetition number of times is the upper limit of the number of repetitions, Simulation means for calculating the value of each node for each iteration and storing the calculated value of each node in the node value information storage unit;
Simulation result display means for displaying the value of each node stored in the node value information storage unit as a simulation result in chronological order in time series;
A time series analysis system characterized by functioning as A terminal device for time series analysis of a model created by an influence diagram,
A node information storage unit for storing node setting information to be set in a node constituting the model in association with a node name and a repetition time in a time series;
A related line information storage unit that stores information of related lines indicating the relationship between the nodes;
A node value information storage unit for storing the value of the node in association with the node name and the information of the repetition times;
A processing unit that performs time series analysis using each information,
The processor is
Information on the nodes constituting the model and information on the related lines input via the input means are acquired, the node information is stored in a node information storage unit, and the information on the related lines is stored in a related line information storage unit. Model information acquisition means stored in
Node setting information acquisition means for acquiring the node setting information input via the input means and storing the node setting information in the node information storage unit;
Obtain information on the upper limit of the number of repetitions input via the input means, and use the information in the node information storage unit and the related line information storage unit until the number of repetitions reaches the upper limit of the number of repetitions. Simulation means for calculating the value of each node for each time and storing the calculated value of each node in the node value information storage unit;
Simulation result display means for displaying the value of each node stored in the node value information storage unit as a simulation result in chronological order in time series;
A terminal device comprising: