Information processing device, information processing method, recording medium, and program

Abstract

A healthcare information processing device is provided. The healthcare information processing device includes an acquisition circuit configured to acquire time series data; a conversion circuit configured to generate a reduced autoregressive model based on the acquired time series data; a creation circuit configured to create a network structure based on the reduced autoregressive model; and a prediction circuit configured to predict a future of an event using the reduced autoregressive model, the event being currently in progress.

The application is claimed as follows: 
     
         1 . A healthcare information processing device comprising:
 an acquisition circuit configured to acquire time series data;   a conversion circuit configured to generate a reduced autoregressive model based on the acquired time series data;   a creation circuit configured to create a network structure based on the reduced autoregressive model; and   a prediction circuit configured to predict a future of an event using the reduced autoregressive model, the event being currently in progress.   
     
     
         2 . The healthcare information processing device according to  claim 1 , wherein the time series data is heart rate data. 
     
     
         3 . The healthcare information processing device according to  claim 1 , wherein the prediction circuit is configured to predict the future of the event including human health. 
     
     
         4 . The healthcare information processing device according to  claim 1 , wherein the reduced autoregressive model satisfies an equation (1): x(t)=a 0 +a 1 x(t−l 1 )+a 2 x (t−l 2 )+ . . . a w x (t−l w )+ε(t), and wherein x(t) is a term at a current time of the event, x(t−l) is a time delay term, a i  (i=0, 1, 2, . . . w) is a parameter, and ε(t) is an approximation error. 
     
     
         5 . The healthcare information processing device according to  claim 1 , wherein nodes of the network are based on time delay terms of the reduced autoregressive model, and the nodes are connected to the term at the current time of the event,
 wherein a direct distance between two nodes of the network is defined by a ratio, and   wherein the ratio is a parameter of a time delay term divided by a square root of a sum of a squares of parameters of the respective terms in the reduced autoregressive model.   
     
     
         6 . The information processing device according to  claim 5 , wherein when two of the nodes are connected to each other via another one or more of the nodes, the creation circuit determines the sum of direct distances between the respective connected nodes as an indirect distance. 
     
     
         7 . The healthcare information processing device according to  claim 5 , wherein the creation circuit is configured to create the network structure by nodes connected to each other with a path having a shortest distance between the direct distance and the indirect distance. 
     
     
         8 . The healthcare information processing device according to  claim 1 , further comprising a retrieval circuit configured to retrieve a homology between different time series data from characteristics of the network structure. 
     
     
         9 . The healthcare information processing device according to  claim 8 , wherein the retrieval circuit is configured to convert the network structure into a matrix, and retrieve the homology of the matrix. 
     
     
         10 . The healthcare information processing device according to  claim 8 , wherein the retrieval circuit is configured to arrange the nodes in a horizontal direction and a vertical direction, and put 1 for the nodes connected to each other and 0 for nodes unconnected to each other to form the matrix. 
     
     
         11 . The healthcare information processing device according to  claim 10 , wherein the matrix comprises a plurality of matrices, and the retrieval circuit is configured to retrieve the homology by a value obtained in such a manner that a difference between the plurality of matrices is calculated after matching the sizes thereof, and the square root of the sum of the squares of elements constituting the difference is divided by the number of the elements. 
     
     
         12 . A healthcare information processing method comprising:
 acquiring time series data;   generating a reduced autoregressive model based on the acquired time series data;   creating a network structure based on the reduced autoregressive model; and   predicting a future of an event using the reduced autoregressive model, the event being currently in progress.   
     
     
         13 . The healthcare information processing method according to  claim 12 , wherein the time series data is heart rate data. 
     
     
         14 . A non-transitory recording medium readable by a computer, the non-transitory recording medium having a healthcare information processing program recorded thereon, the healthcare information processing program causing the computer to execute:
 an acquisition step of acquiring time series data;   a generation step of generating a reduced autoregressive model based on the acquired time series data;   a creation step of creating a network structure based on the reduced autoregressive model; and   a prediction step of predicting a future of an event using the reduced autoregressive model, the event being currently in progress.   
     
     
         15 . The non-transitory recording medium according to  claim 14 , wherein the time series data is heart rate data.