Policy introduced effect prediction apparatus, and policy introduced effect prediction method

A non-transitory computer-readable recording medium having stored therein a program for causing a computer to perform a policy introduced effect prediction process comprising:

• receiving, from a terminal via a network, a simulation request, including an objective variable name, an explanatory variable name, and an expected value of an explanatory variable in response to introduction of a policy, which is input by a user on the terminal;
• calculating at least one calculation model based on first difference data calculated based on time series data of an objective variable, which is a variable indicating an effect of introducing the policy, and second difference data calculated based on time series data of the explanatory variable, which is a variable highly related to the objective variable;
• calculating a difference amount of the objective variable by inputting a difference amount of the explanatory variable for introducing a policy into the calculated calculation model;
• calculating a predicted value of the objective variable in response to introduction of the policy by adding the calculated difference amount of the objective variable to a predicted value of the objective variable in the absence of introduction of the policy;
• and
• transmitting, to the terminal via the network, a simulation result including the predicted value of the objective variable in response to introduction of the policy, to display on the terminal a display screen representing a difference between the predicted value of the objective variable in the absence of introduction of the policy and the predicted value of the objective variable in response to introduction of the policy, and enable the user to quantitatively recognize the effect of introducing the policy,
• wherein the calculating of the at least one calculation model includes:
• determining whether there is a positive correlation or a negative correlation between the first difference data calculated based on the time series data of the objective variable received from a storage device via the network and the second difference data calculated based on the time series data of the explanatory variable received from the storage device via the network;
• in response to determining that there is a positive correlation between the first difference data and the second difference data, extracting difference data with same sign from among the first difference data and the second difference data and calculating a first calculation model having a positive correlation based on the extracted difference data with the same sign;
• and
• in response to determining that there is a negative correlation between the first difference data and the second difference data, extracting difference data with different signs from among the first difference data and the second difference data and calculating a second calculation model having a negative correlation based on the extracted difference data with the different signs.

A policy introduced effect prediction system comprising:

• a policy introduced effect prediction apparatus;
• a terminal configured to communicate with the policy introduced effect prediction apparatus;
• and
• a storage device configured to store time series data, wherein
• the policy introduced effect prediction apparatus includes a memory storing a set of instructions, and one or more processors programed to execute the set of instructions to perform a policy introduced effect prediction process including:
• receiving, from a terminal via a network, a simulation request, including an objective variable name, an explanatory variable name, and an expected value of an explanatory variable in response to introduction of a policy, which is input by a user on the terminal;
• calculating at least one calculation model based on first difference data calculated based on time series data of an objective variable, which is a variable indicating an effect of introducing the policy, and second difference data calculated based on time series data of the explanatory variable, which is a variable highly related to the objective variable;
• calculating a difference amount of the objective variable by inputting a difference amount of the explanatory variable for introducing a policy into the calculated calculation model;
• calculating a predicted value of the objective variable in response to introduction of the policy by adding the calculated difference amount of the objective variable to a predicted value of the objective variable in the absence of introduction of the policy;
• and
• transmitting, to the terminal via the network, a simulation result including the predicted value of the objective variable in response to introduction of the policy, to display on the terminal a display screen representing a difference between the predicted value of the objective variable in the absence of introduction of the policy and the predicted value of the objective variable in response to introduction of the policy, and enable the user to quantitatively recognize the effect of introducing the policy,
• wherein the calculating of the at least one calculation model includes:
• determining whether there is a positive correlation or a negative correlation between the first difference data calculated based on the time series data of the objective variable received from the storage device via the network and the second difference data calculated based on the time series data of the explanatory variable received from the storage device via the network;
• in response to determining that there is a positive correlation between the first difference data and the second difference data, extracting difference data with same sign from among the first difference data and the second difference data and calculating a first calculation model having a positive correlation based on the extracted difference data with the same sign;
• and
• in response to determining that there is a negative correlation between the first difference data and the second difference data, extracting difference data with different signs from among the first difference data and the second difference data and calculating a second calculation model having a negative correlation based on the extracted difference data with the different signs.