JP7489731B2 - Clustering processing device, clustering processing method, program, and information processing device - Google Patents

Description

The present invention relates to a clustering processing device, a clustering processing method, a program, and an information processing device, and more particularly to a clustering processing device, a clustering processing method, a program, and an information processing device that sequentially inputs input vectors belonging to an arbitrary class and learns the input distribution structure of the input vectors.

In recent years, a method has been known in which various input data are subjected to supervised learning to classify the input data. In this case, it is necessary to prepare training data in which appropriate labels have been assigned to the data in advance. To create such training data, input data that has not been assigned labels for creating training data is prepared, and clustering is performed by unsupervised learning using a clustering method such as the k-means method to create training data in which labels are assigned to each cluster.

As an efficient learning method for creating such training data, for example, a method called Self-Organizing Incremental Neural Network (SOINN), which is a learning method that grows neurons as needed during learning, has been proposed. SOINN has many advantages, such as being able to learn non-stationary inputs by autonomously managing the number of nodes, and being able to extract an appropriate number of classes and topological structure even for classes with complex distribution shapes. As an example of an application of SOINN, for example in pattern recognition, after learning classes of hiragana characters, it is possible to additionally learn classes of katakana characters, etc.

As an example of such SOINN, a method called E-SOINN (Enhanced SOINN) has been proposed (Patent Document 1). E-SOINN allows for online additional learning, which adds learning at any time, and has the advantage of being more efficient than batch learning. Therefore, E-SOINN allows additional learning even when the learning environment changes to a new environment. E-SOINN also has the advantage of being highly resistant to noise in input data.

However, in SOINNs including E-SOINN, it is difficult to insert new nodes into the network, which makes it difficult to accurately express the structure of the input data, and the learning results differ depending on the input order of the input data. To solve these problems, a method called LB-SOINN (Load Balance Self-Organizing Incremental Neural Network) has been proposed (Patent Document 2). LB-SOINN treats the load of nodes in the network as node learning time, detects nodes with large node learning times, and generates a new node on the edge connecting the detected node to an adjacent node, with a weight vector determined based on the weight vector of the detected node. This alleviates the increase in the learning time of the detected node, and by generating a new node in its vicinity, the structure of the input data can be learned more accurately.

JP 2008-217246 A JP 2014-164396 A

Burr Settles, “Active Learning Literature Survey”, January 9, 2009, Computer Sciences Technical Report 1648, University of Wisconsin-Madison

However, even with these methods, it may not be possible to assign labels to all nodes after clustering. In such cases, an operator must refer to the results of clustering and manually assign labels that they deem appropriate to nodes that have not been labeled. This makes it time-consuming to create training data. In such cases, it is possible that the training data obtained will vary depending on the operator performing the work.

It is also possible to apply other methods such as active learning (Non-Patent Document 1) to assign labels to unlabeled nodes. However, even in this case, if there are still unlabeled nodes remaining, it is still necessary to manually assign labels. In addition, even with this method, training data must be created first, so manual labeling is also required here.

Furthermore, if labeling is done manually, if the amount of data is excessively large, it becomes difficult to label in the first place.

As such, it is difficult to create training data quickly and automatically using general training data creation methods. Therefore, a method is needed that can assign appropriate labels to all nodes in unsupervised learning of input data.

The present invention has been made in consideration of the above circumstances, and aims to provide a clustering processing device, a clustering processing method, a program, and an information processing device that are capable of assigning appropriate labels to all nodes in unsupervised learning of input data.

A clustering processing device according to one embodiment of the present invention includes an initial clustering processing unit that clusters input data consisting of a plurality of unlabeled nodes described by a multidimensional vector, and acquires clustering intermediate data in which the nodes belonging to the clusters are labeled, and an additional clustering processing unit that creates clustering result data in which the unlabeled nodes included in the clustering intermediate data are assigned the same label as that assigned to a node that belongs to any of the clusters included in the clustering intermediate data and is located the shortest distance from the unlabeled node. This makes it possible to create clustering result data in which the nodes not assigned a label in the initial clustering process are assigned the same label as that assigned to the node that belongs to any of the clusters and is located the shortest distance from the unlabeled node.

In one embodiment of the present invention, the clustering processing device is the above-mentioned clustering processing device, in which the additional clustering processing unit includes a node selection unit that selects one unlabeled node that does not belong to any of the clusters from the nodes included in the clustering intermediate data, a distance calculation unit that calculates the distance between the selected unlabeled node and all nodes that belong to the cluster, an affiliated cluster determination unit that identifies the shortest distance node from all nodes that belong to the cluster based on the calculated distance, and a label assignment unit that assigns the same label as the label assigned to the shortest distance node to the selected unlabeled node. This makes it possible to calculate the shortest distance between a node that was not assigned a label in the initial clustering process and a node that belongs to a cluster, and assign the same label as the shortest distance node to the node that was not assigned a label.

In a clustering processing device according to one embodiment of the present invention, in the clustering processing device described above, the additional clustering processing unit further includes a progress determination unit that determines whether or not there is a node that is not labeled in the clustering intermediate data, and the additional clustering processing unit repeats processing by the node selection unit, the distance calculation unit, the belonging cluster determination unit, and the label assignment unit until the progress determination unit determines that there is or is not a node that is not labeled in the clustering intermediate data. This makes it possible to assign labels to pairs of all nodes that were not assigned labels in the initial clustering process.

A clustering processing method according to one embodiment of the present invention clusters input data consisting of a plurality of unlabeled nodes described by a multidimensional vector, obtains clustering intermediate data in which labels are assigned to nodes belonging to the clusters, and creates clustering result data in which unlabeled nodes included in the clustering intermediate data are assigned the same label as that assigned to a node that belongs to any of the clusters included in the clustering intermediate data and is located the shortest distance from the unlabeled node. This makes it possible to create clustering result data in which nodes that were not assigned a label in the initial clustering process are assigned the same label as that assigned to a node that belongs to any of the clusters and is located the shortest distance away.

A program according to one embodiment of the present invention causes a computer to execute the following processes: clustering input data consisting of a plurality of unlabeled nodes described by a multidimensional vector, acquiring clustering intermediate data in which nodes belonging to the clusters are labeled; and creating clustering result data in which unlabeled nodes included in the clustering intermediate data are assigned the same label as that assigned to a node that belongs to any of the clusters included in the clustering intermediate data and is located the shortest distance from the unlabeled node. This makes it possible to create clustering result data in which nodes that were not assigned a label in the initial clustering process are assigned the same label as that assigned to a node that belongs to any of the clusters and is located the shortest distance away.

An information processing device according to an embodiment of the present invention includes a teacher data creation unit that creates teacher data by performing clustering processing on teacher data creation input data consisting of nodes described by multidimensional vectors that are not assigned multiple labels, a supervised learning unit that assigns labels to nodes of learning target input data consisting of nodes described by multidimensional vectors that are not assigned multiple labels based on the teacher data, and a display unit that displays the processing results by the supervised learning unit. The teacher data creation unit includes an initial clustering processing unit that clusters input data consisting of nodes described by multidimensional vectors that are not assigned multiple labels, and obtains clustering intermediate data in which labels are assigned to nodes belonging to clusters, and an additional clustering processing unit that creates clustering result data in which unlabeled nodes included in the clustering intermediate data are assigned the same label as that assigned to a node that is the shortest distance from the unlabeled node among nodes included in any of the clusters included in the clustering intermediate data. This makes it possible to create clustering result data in which nodes that were not assigned labels in the initial clustering processing are assigned the same label as that assigned to a node that is the shortest distance from any of the clusters.

According to the present invention, it is possible to provide a clustering processing device, a clustering processing method, a program, and an information processing device that can assign appropriate labels to all nodes after clustering in unsupervised learning of input data.

FIG. 1 is a diagram illustrating an example of a system configuration for realizing an information processing device according to a first embodiment. 1 is a diagram illustrating a basic configuration of an information processing device according to a first embodiment; FIG. 2 is a diagram showing a more detailed configuration of the information processing device according to the first embodiment. FIG. 13 is a diagram illustrating another example of the configuration of the information processing device according to the first embodiment. 2 is a diagram illustrating a basic configuration of a teacher data creation unit according to the first embodiment; FIG. FIG. 1 is a diagram showing an example of input data IN1 for creating teacher data used to create teacher data. 11 is a diagram showing an example of clustering intermediate data acquired by the clustering process of the initial clustering processor; FIG. 2 is a diagram showing a more detailed configuration of a teacher data creation unit according to the first embodiment; FIG. 11 is a flowchart of a clustering process performed by a teacher data creation unit according to the first embodiment. 2 shows the node selected by the node selection unit. FIG. 13 is a diagram showing an example of training data D _TCH that is created by automatically assigning labels to all nodes.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In each drawing, the same elements are given the same reference numerals, and duplicate explanations will be omitted as necessary.

First embodiment
FIG. 1 is a diagram showing an example of a system configuration for realizing an information processing device according to the first embodiment. The information processing device 1000 can be realized by a computer 10 such as a dedicated computer or a personal computer (PC). However, the computer does not need to be physically single, and may be multiple when performing distributed processing. As shown in FIG. 1, the computer 10 has a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, and a RAM (Random Access Memory) 13, which are connected to each other via a bus 14. Note that an OS software for operating the computer will not be described, but it is assumed that the computer that constructs this information processing device also has such software.

An input/output interface 15 is also connected to the bus 14. The input/output interface 15 is connected to, for example, an input unit 16 including a keyboard, mouse, sensor, etc., a display including a CRT, LCD, etc., an output unit 17 including headphones, speakers, etc., a storage unit 18 including a hard disk, etc., and a communication unit 19 including a modem, terminal adapter, etc.

The CPU 11 executes various processes according to various programs stored in the ROM 12 or various programs loaded from the storage unit 18 to the RAM 13, and in this embodiment, for example, the processes of each part of the information processing device 1000 described later. A GPU (Graphics Processing Unit) may be provided separately from the CPU 11, and similarly to the CPU 11, executes various processes according to various programs stored in the ROM 12 or various programs loaded from the storage unit 18 to the RAM 13, and in this embodiment, for example, the processes of each part of the information processing device 1000 described later. Note that the GPU is suitable for applications in which standard processes are performed in parallel, and by applying it to processes in a neural network described later, it is possible to improve the processing speed compared to the CPU 11. The RAM 13 also appropriately stores data necessary for the CPU 11 and the GPU 21 to execute various processes.

The communication unit 19 performs communication processing, for example, via the Internet (not shown), transmits data provided by the CPU 11, and outputs data received from a communication partner to the CPU 11, RAM 13, and storage unit 18. The storage unit 18 communicates with the CPU 11 and stores and erases information. The communication unit 19 also performs communication processing of analog or digital signals with other devices.

The input/output interface 15 is also connected to a drive 20 as required, and, for example, a magnetic disk 20A, an optical disk 20B, a flexible disk 20C, or a semiconductor memory 20D is appropriately attached, and computer programs read from them are installed in the memory unit 18 as required.

Next, each process in the information processing device 1000 according to this embodiment will be described. The information processing device 1000 receives as input a non-hierarchical neural network in which nodes described by n-dimensional vectors (n is an integer equal to or greater than 1) are arranged. The neural network is stored in a storage unit such as the RAM 13.

The neural network in this embodiment is a self-replicating neural network that inputs an input vector into the neural network and automatically increases the number of nodes placed in the neural network based on the input vector. By using a self-replicating neural network, the number of nodes can be automatically increased.

The neural network in this embodiment has a non-hierarchical structure. By adopting a non-hierarchical structure, additional learning can be performed without specifying the timing to start learning in other layers. In other words, additional learning can be performed online.

The input data is input as an n-dimensional input vector. For example, the input vector is stored in a temporary storage unit (e.g., RAM 13) and is input sequentially to the neural network stored in the temporary storage unit.

The specific configuration of the information processing device 1000 according to the first embodiment will be described below. The information processing device 1000 is first given input data for creating teacher data used to create teacher data, and the teacher data is created based on this. The information processing device 1000 then uses the created teacher data to assign labels to data (nodes) included in the input data for separately provided learning input data, using the teacher data.

Figure 2 shows a schematic diagram of the basic configuration of the information processing device 1000 according to the first embodiment. Figure 3 shows a more detailed diagram of the configuration of the information processing device 1000 according to the first embodiment. The information processing device 1000 has at least a teacher data creation unit 100, and may further have a supervised learning unit 110 and a display unit 120. In this example, input data IN1 for creating teacher data and learning target input data IN2 to be subjected to supervised learning are provided to the teacher data creation unit 100 and the supervised learning unit 110, respectively, from outside the information processing device 1000. As described above, the input data IN1 for creating teacher data and learning target input data IN2 to be subjected to supervised learning are data including multiple nodes described by multidimensional vectors.

The input data IN1 for creating teacher data and the learning target input data IN2 may be stored in a storage unit provided in the information processing device 1000. FIG. 4 shows another configuration example of the information processing device 1000 according to the first embodiment. In this example, the information processing device 1000 further includes a storage unit 130. The input data IN1 for creating teacher data and the learning target input data IN2 are appropriately stored in the storage unit 130, and the teacher data creation unit 100 and the teacher-assisted learning unit 110 may read the input data IN1 for creating teacher data and the learning target input data IN2 from the storage unit 130 as necessary. The input data IN1 for creating teacher data and the learning target input data IN2 may be stored in the storage unit 130 in advance or at any time by the operator of the information processing device 1000. The storage unit 130 corresponds to, for example, one or both of the ROM 12 and the storage unit 18 shown in FIG. 1.

The teacher data creation unit 100 is configured to perform clustering on the input data IN1 for creating teacher data using an unsupervised learning method such as the k-means method or the above-mentioned SOINN method to create clustering intermediate data, and to create teacher data D _TCH by further assigning labels to unlabeled nodes included in the clustering intermediate data.

In other words, the teacher data creation unit 100 is configured to perform a clustering process that can automatically assign labels to all nodes belonging to a cluster after the first clustering process.

Note that the application of the clustering process performed by the teacher data creation unit 100 is not limited to the creation of teacher data, but can be applied to clustering processes of various input data. Therefore, the teacher data creation unit 100 is also simply referred to as a clustering processing device.

Figure 5 shows a schematic diagram of the basic configuration of the teacher data creation unit 100. The teacher data creation unit 100 has an initial clustering processing unit 100A and an additional clustering processing unit 100B.

The initial clustering processing unit 100A is configured to perform clustering on the input data IN1 for creating teacher data using an unsupervised learning method such as the k-means method or the above-mentioned SOINN method, and to create clustering intermediate data D _{INT_L} in which labels are assigned to the nodes.

Figure 6 shows an example of the input data IN1 for creating teacher data used to create teacher data. In this example, data in which unlabeled nodes (shown as white circles) are distributed on a two-dimensional plane is used as the input data IN1 for creating teacher data.

7 shows an example of clustering intermediate data D _{INT_L} acquired by the initial clustering process of the initial clustering processing unit 100A. The clustering intermediate data D _{INT_L} in FIG. 7 includes, for example, four clusters C1 to C4, and a plurality of nodes belong to each of the clusters C1 to C4. Note that in the clustering intermediate data D _{INT_L} , not all of the included nodes belong to any cluster, and there are nodes that do not belong to any cluster and that have not been assigned any label. In FIG. 7, nodes that do not belong to any cluster and that have not been assigned any label are displayed as unlabeled nodes with white circles.

As described above, the clustering intermediate data D _{INT_L} includes unlabeled nodes, and further clustering processing is performed in the subsequent additional clustering processing unit 100B. Therefore, for the sake of distinction, the processing entity here is referred to as the initial clustering processing unit 100A, and the clustering processing performed by the initial clustering processing unit 100A is referred to as the initial clustering processing.

The additional clustering processing unit 100B further performs processing to assign labels to unlabeled nodes included in the clustering intermediate data D _{INT_L} , and creates teacher data D _TCH , which is clustering result data.

Here, to distinguish it from the initial clustering process, the processing entity here will be referred to as the additional clustering processing unit 100B, and the clustering process performed by the additional clustering processing unit 100B will be referred to as the additional clustering process.

Next, the configuration and operation of the teacher data creation unit 100 will be described in more detail. Figure 8 shows the configuration of the teacher data creation unit 100 according to the first embodiment in more detail. The initial clustering processing unit 100A has a data acquisition unit 101, a clustering processing unit 102, and a first label assignment unit 103. The additional clustering processing unit 100B has a node selection unit 104, a distance calculation unit 105, an affiliated cluster determination unit 106, a second label assignment unit 107, and a progress determination unit 108.

9 is a flowchart showing the clustering process performed by the teacher data creation unit 100. The clustering process performed by the teacher data creation unit 100 goes through the following steps S1 to S8 to create teacher data _{D_TCH} , which is clustering result data.

Step S1
The data acquisition unit 101 acquires input data IN1 for creating teacher data from outside the information processing device 1000 or from the memory unit 130.

Step S2
The clustering processing unit 102 performs initial clustering processing on the input data IN1 for creating teacher data using an unsupervised learning method such as the k-means method or the above-mentioned SOINN method, and obtains clustering intermediate data D _INT , which is the clustering result. As described above, the clustering intermediate data D _INT at this stage also includes nodes that do not belong to any cluster.

Step S3
The first label assignment unit 103 assigns a label to a node that has come to belong to any one of the clusters C1 to C4. For example, the first label assignment unit 103 may assign labels "C1", "C2", "C3", and "C4" to the nodes that have come to belong to each of the clusters C1 to C4. However, the labels assigned to the nodes that have come to belong to any one of the clusters C1 to C4 are merely examples, and other appropriate labels may be assigned as necessary. The first label assignment unit 103 outputs the clustering intermediate data D _{INT_L} to which labels have been assigned. As described above, the clustering intermediate data D _{INT_L} at this stage also includes unlabeled nodes.

Step S4
The node selection unit 104 selects one unlabeled node included in the labeled clustering intermediate data D _{INT_L} , and outputs the selection result SEL. Fig. 10 shows the nodes selected by the node selection unit 104. In this example, it is assumed that the unlabeled node NS in Fig. 10 is selected.

Step S5
Based on the selection result SEL, the distance calculation unit 105 calculates the distance between the selected node NS and each node belonging to the clusters C1 to C4. The distance calculation unit 105 outputs distance information DIS indicating the calculated distance. As the distance measure applied here, various distance measures such as cosine distance, Euclidean distance, Mahalanobis distance, Manhattan distance, and fractional distance can be used.

Step S6
The cluster determining unit 106 detects the shortest distance among the distances calculated based on the distance information DIS, and identifies the node that is at the shortest distance DIS _MIN from the selected node NS among the nodes that belong to clusters C1 to C4. Here, as an example, it is assumed that one of the nodes that belong to cluster C4 is the shortest distance node NN. The cluster determining unit 106 outputs node identification information ND that indicates the identified node.

Step S7
Based on the node identification information ND, the second label assignment unit 107 assigns to the unlabeled node NS the same label as the label assigned to the node identified by the cluster determination unit 106. For example, if the label "C4" is assigned to the shortest distance node NN belonging to cluster C4, the cluster determination unit 106 assigns the same label "C4" to the unlabeled node NS. As a result, the unlabeled node NS to which the label has been assigned becomes a node belonging to cluster C4.

Step S8
The progress determination unit 108 determines whether the clustering intermediate data D _{INT_L} after labels are assigned includes an unlabeled node. If the clustering intermediate data D _{INT_L} includes an unlabeled node, the process returns to step S4. As a result, the processes of steps S4 to S7 are repeated until all of the unlabeled nodes included in the clustering intermediate data D _{INT_L} belong to any of the clusters including the node with the shortest distance.

On the other hand, if the labeled clustering intermediate data D _{INT_L} does not include any unlabeled nodes, the progress determination unit 108 ends the clustering process and outputs the latest clustering intermediate data D _{INT_L} as teacher data D _TCH .

By the above processing shown in steps S1 to S8, even if unlabeled nodes remain in the initial clustering processing, labels can be automatically assigned to all nodes. Figure 11 shows an example of teacher data D _TCH created by automatically assigning labels to all nodes. As shown in Figure 11, there are no unlabeled nodes in the teacher data D _TCH , and it can be seen that each node belongs to one of the clusters C1 to C4.

The supervised learning unit 110 assigns labels to the separately provided learning target input data IN2 based on the teacher data _DTCH created as described above.

The display unit 120 can appropriately display intermediate and final results of clustering by the teacher data creation unit 100, processing results by the supervised learning unit 110, etc.

As described above, with this configuration, in unsupervised learning of input data, appropriate labels can be automatically assigned to all nodes after clustering.

This makes it possible to automatically and quickly create training data without manual intervention. It also makes it possible to automatically perform a series of processes, from creating training data to unsupervised learning using the training data and displaying the learning results.

Other embodiments The present invention is not limited to the above-mentioned embodiments, and can be modified as appropriate without departing from the spirit of the present invention. For example, regarding the distance measure, when performing online additional learning, sample data cannot be obtained in advance, so it is not possible to analyze the number of dimensions of the input vector in advance to determine which distance measure is effective. For this reason, as explained using equation (14) in Patent Document 2, a new distance measure that represents the distance between two nodes by combining different distance measures may be introduced. For example, as shown in equation (17) derived using equations (14) to (16) in Patent Document 2, a new distance measure that combines Euclidean distance and cosine distance may be used.

In addition, although the distance measure has been described as a combination of Euclidean distance and cosine distance, the present invention is not limited to this and may be combined with other distance measures (e.g., cosine distance, Manhattan distance, fractional distance). Furthermore, the present invention is not limited to distance measures that are effective in high-dimensional spaces and may be combined with other distance measures depending on the problem to be studied.

In the above embodiment, the present invention has been described mainly as a hardware configuration, but the present invention is not limited to this, and any process can be realized by having a CPU (Central Processing Unit) execute a computer program. In this case, the computer program can be stored using various types of non-transitory computer readable medium and supplied to the computer. Non-transitory computer readable medium includes various types of tangible storage medium. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R/Ws, and semiconductor memories (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (random access memories)). The program may also be provided to the computer by various types of transient computer-readable media. Examples of the transient computer-readable medium include electric signals, optical signals, and electromagnetic waves. The transient computer-readable medium can provide the program to the computer via a wired communication path such as an electric wire or optical fiber, or via a wireless communication path.

The present invention has been described above with reference to the embodiment, but the present invention is not limited to the above. Various modifications that can be understood by a person skilled in the art can be made to the configuration and details of the present invention within the scope of the invention.

This application claims priority based on Japanese Patent Application No. 2021-6621, filed on January 19, 2021, the disclosure of which is incorporated herein in its entirety.

10 Computer 11 CPU
12 ROM
13 RAM
14 Bus 15 Input/Output Interface 16 Input Unit 17 Output Unit 18 Memory Unit 19 Communication Unit 20 Drive 20A Magnetic Disk 20B Optical Disk 20C Flexible Disk 20D Semiconductor Memory 100 Teacher Data Creation Unit 100A Initial Clustering Processing Unit 100B Additional Clustering Processing Unit 110 Supervised Learning Unit 120 Display Unit 130 Memory Unit 101 Data Acquisition Unit 102 Clustering Processing Unit 103 First Label Assignment Unit 104 Node Selection Unit 105 Distance Calculation Unit 106 Belonging Cluster Determination Unit 107 Second Label Assignment Unit 108 Progress Judgment Unit 1000 Information Processing Device C1 to C4 Clusters D _INT , D _{INT_L} Clustering Intermediate Data D _TCH Teacher Data IN1 Input Data for Creating Teacher Data IN2 Learning Target Input Data

Claims (6)

an initial clustering processing unit that clusters input data consisting of a plurality of unlabeled nodes described by a multidimensional vector, and obtains clustering intermediate data in which labels are assigned to nodes belonging to the clusters;
an additional clustering processing unit that creates clustering result data in which, for nodes included in the clustering intermediate data that are not labeled, the same label as that assigned to a node that is located at the shortest distance from the unlabeled node among nodes that belong to any of the clusters included in the clustering intermediate data is assigned to the nodes included in the clustering intermediate data that are not labeled;
Clustering processing unit. The additional clustering processing unit includes:
a node selection unit that selects, from the nodes included in the clustering intermediate data, one unlabeled node that does not belong to any of the clusters;
a distance calculation unit that calculates a distance between the selected unlabeled node and all nodes that belong to the cluster;
a cluster determining unit that identifies a node that is closest to the selected one unlabeled node from among all nodes that belong to the cluster based on the calculated distance;
a label assignment unit that assigns the same label as the label assigned to the shortest distance node to the selected one unlabeled node,
The clustering processing device according to claim 1 . the additional clustering processing unit further includes a progress determination unit that determines whether or not there is an unlabeled node in the clustering intermediate data;
the additional clustering processing unit repeats the processes by the node selection unit, the distance calculation unit, the belonging cluster determination unit, and the label assignment unit until the progress determination unit determines that there is no node not having a label in the clustering intermediate data.
The clustering processing device according to claim 2 . An initial clustering processing unit clusters input data consisting of a plurality of unlabeled nodes described by a multidimensional vector, and obtains clustering intermediate data in which labels are assigned to nodes belonging to the clusters;
creating clustering result data in which an additional clustering processing unit assigns to an unlabeled node included in the clustering intermediate data the same label as that assigned to a node that is located at the shortest distance from the unlabeled node among nodes that belong to any one of the clusters included in the clustering intermediate data;
Clustering processing method. A process of clustering input data consisting of a plurality of unlabeled nodes described by a multidimensional vector, and acquiring intermediate clustering data in which labels are assigned to nodes belonging to the clusters;
creating clustering result data in which, for unlabeled nodes included in the clustering intermediate data, the same label as that assigned to a node that is located the shortest distance from the unlabeled node among nodes that belong to any one of the clusters included in the clustering intermediate data is assigned to the unlabeled node;
program. a teacher data creation unit that creates teacher data by performing a clustering process on teacher data creation input data consisting of a plurality of unlabeled nodes described by a multidimensional vector;
a supervised learning unit that assigns labels to nodes of learning target input data that includes a plurality of unlabeled nodes described by a multidimensional vector based on the teaching data;
a display unit that displays a processing result by the supervised learning unit,
The teacher data creation unit is
an initial clustering processing unit that clusters input data consisting of a plurality of unlabeled nodes described by a multidimensional vector, and obtains clustering intermediate data in which labels are assigned to nodes belonging to the clusters;
an additional clustering processing unit that creates clustering result data in which, for nodes included in the clustering intermediate data that are not labeled, the same label as that assigned to a node that is located at the shortest distance from the unlabeled node among nodes that belong to any of the clusters included in the clustering intermediate data is assigned to the nodes included in the clustering intermediate data that are not labeled;
Information processing device.

Images