Information processing device, system, information processing method, and program

Abstract

To improve convenience for an observer by simultaneously classifying multiple kinds of normal values and abnormal values related to water level data.SOLUTION: An information processing apparatus includes: an acquisition unit which acquires time-series data on water levels measured by a water level gauge; a generation unit which generates multiple pieces of vector data including time-series values of the water levels and labels indicating a state of the water levels on the basis of the time-series data; a classification unit which classifies the multiple pieces of vector data into a plurality of nodes using a self-organizing map (SOM); and a determination unit which determines, using a majority voting method, representative labels representing each of the nodes, on the basis of the labels assigned to the multiple pieces of vector data classified into the nodes.SELECTED DRAWING: Figure 1

Claims

1. An acquisition unit that acquires time-series data of water level measured by a water level gauge, A generation unit generates a plurality of vector data, including the time-series values ​​of the water level and labels representing the state of the water level, based on the aforementioned time-series data. A classification unit that classifies the aforementioned multiple vector data into multiple nodes using a Self-Organizing Map (SOM), The system includes a determination unit that determines a representative label representing each of the plurality of nodes using a majority voting method, based on the labels assigned to the plurality of vector data classified into each of the plurality of nodes, The generation unit generates vector data that includes a vector whose components are time-series values ​​of adjacent water levels measured during the period from a predetermined number of measurement steps prior to the target time point for assigning the label to the measurement step at the target time point. Information processing device.

2. The system further includes a display control unit that causes the representative label determined for each of the plurality of nodes to be displayed on the user terminal. The information processing apparatus according to claim 1.

3. An acquisition unit that acquires time-series data of the water level measured by a water level gauge, A generation unit generates a plurality of vector data, including the time-series values ​​of the water level and labels representing the state of the water level, based on the aforementioned time-series data. A classification unit that classifies the aforementioned multiple vector data into multiple nodes using a Self-Organizing Map (SOM), The system includes a determination unit that determines a representative label representing each of the plurality of nodes using a majority voting method, based on the labels assigned to the plurality of vector data classified into each of the plurality of nodes, The labels indicating the state of the water level include a label indicating that the water level is normal, a label indicating that a slide displacement has occurred in the water level, a label indicating that spike noise has occurred in the water level, and a label indicating that a flood event has occurred in the water level. Information processing device.

4. The majority voting method determines the representative label from among the labels assigned to the plurality of vector data classified into each of the plurality of nodes, the label with the largest number of votes. The information processing apparatus according to claim 1.

5. An information processing device according to any one of claims 1 to 4, The water level gauge and, The system includes a transmitter that transmits the time-series data measured by the water level gauge to the information processing device. system.

6. Computers Time-series data of water levels measured by water level gauges is acquired. Based on the aforementioned time-series data, a plurality of vector data sets are generated, each including the time-series value of the water level and a label representing the state of the water level. The aforementioned multiple vector data are classified into multiple nodes using a Self-Organizing Map (SOM). Based on the labels assigned to the plurality of vector data classified into each of the plurality of nodes, a representative label representing each of the plurality of nodes is determined using a majority vote method. The generation process generates vector data that includes a vector whose components are time-series values ​​of adjacent water levels measured during the period from a predetermined number of measurement steps prior to the target time point for assigning the label to the measurement step at the target time point. Information processing methods.

7. On the computer, Time-series data of water levels measured by a water level gauge is acquired. Based on the aforementioned time-series data, a plurality of vector data sets are generated, each including the time-series value of the water level and a label representing the state of the water level. The aforementioned multiple vector data are classified into multiple nodes using a Self-Organizing Map (SOM). Based on the labels assigned to the plurality of vector data classified into each of the plurality of nodes, a representative label representing each of the plurality of nodes is determined using a majority vote method. The generation process generates vector data that includes a vector whose components are time-series values ​​of adjacent water levels measured during the period from a predetermined number of measurement steps prior to the target time point for assigning the label to the measurement step at the target time point. program.

Images