18503618. INFORMATION PROCESSING APPARATUS, INFORMATION PROCESSING METHOD, AND STORAGE MEDIUM simplified abstract (Rakuten Group, Inc.)
Contents
INFORMATION PROCESSING APPARATUS, INFORMATION PROCESSING METHOD, AND STORAGE MEDIUM
Organization Name
Inventor(s)
Som Subhra Ghosh of Tokyo (JP)
INFORMATION PROCESSING APPARATUS, INFORMATION PROCESSING METHOD, AND STORAGE MEDIUM - A simplified explanation of the abstract
This abstract first appeared for US patent application 18503618 titled 'INFORMATION PROCESSING APPARATUS, INFORMATION PROCESSING METHOD, AND STORAGE MEDIUM
The patent application describes an information processing apparatus that can classify delivery destinations into groups using a graph convolutional neural network without the need for pre-prepared training data.
- The apparatus includes a memory storing a program and at least one processor that executes the program.
- The processor performs unsupervised learning to train a graph convolutional neural network using an adjacency matrix and a feature matrix of delivery destinations.
- The learning unit uses a loss function to minimize the distance between delivery destinations in the same group and the difference in features within the same group.
- The apparatus outputs information about the group to which the delivery destinations belong based on the trained neural network.
Potential Applications: - Logistics and supply chain management for efficient routing and delivery grouping. - Social network analysis for community detection and recommendation systems.
Problems Solved: - Automating the process of graph partitioning without the need for labeled training data. - Improving the efficiency of grouping and classification tasks in various industries.
Benefits: - Saves time and resources by eliminating the need for manual data preparation. - Enhances accuracy and scalability of graph partitioning tasks.
Commercial Applications: Title: "Automated Graph Partitioning for Efficient Delivery Routing" This technology can be utilized in logistics companies to optimize delivery routes, in social media platforms for targeted content delivery, and in research institutions for network analysis.
Prior Art: Further research can be conducted in the fields of machine learning, graph theory, and neural networks to explore similar approaches to unsupervised graph partitioning.
Frequently Updated Research: Stay updated on advancements in graph convolutional neural networks, unsupervised learning techniques, and applications in various industries.
Questions about Graph Convolutional Neural Networks: 1. How do graph convolutional neural networks differ from traditional neural networks in handling graph data? Graph convolutional neural networks are specifically designed to operate on graph-structured data, capturing the relationships between nodes in a graph.
2. What are some challenges in training graph convolutional neural networks for unsupervised learning tasks? Training graph convolutional neural networks in an unsupervised manner can be challenging due to the complexity of capturing meaningful patterns in the graph structure without labeled data.
Original Abstract Submitted
To enable graph partitioning using a graph convolutional neural network without preparing training data, an information processing apparatus is configured to classify a plurality of delivery destinations into a plurality of groups, and includes: a memory storing a program; and at least one processor that, by executing the program stored in the memory, is configured to: perform unsupervised learning to train a graph convolutional neural network, which is determined using an adjacency matrix indicating a connection relationship of the plurality of delivery destinations, and receives as input a feature matrix indicating a feature of the plurality of delivery destinations, the learning unit performing unsupervised learning using a first loss function defined such that the smaller a value for distance between delivery destinations belonging to a same group and the smaller a difference in features between delivery destinations belonging to a same group, the less a loss; and output information about a group to which the plurality of delivery destinations belongs, the information being obtained by inputting the feature matrix into the graph convolutional neural network trained by the learning unit.
