18234883. COMPUTER-READABLE RECORDING MEDIUM STORING MACHINE LEARNING PROGRAM, INFORMATION PROCESSING DEVICE, AND MACHINE LEARNING METHOD simplified abstract (Fujitsu Limited)

COMPUTER-READABLE RECORDING MEDIUM STORING MACHINE LEARNING PROGRAM, INFORMATION PROCESSING DEVICE, AND MACHINE LEARNING METHOD

Organization Name

Fujitsu Limited

Inventor(s)

Yuichi Kamata of Isehara (JP)

COMPUTER-READABLE RECORDING MEDIUM STORING MACHINE LEARNING PROGRAM, INFORMATION PROCESSING DEVICE, AND MACHINE LEARNING METHOD - A simplified explanation of the abstract

This abstract first appeared for US patent application 18234883 titled 'COMPUTER-READABLE RECORDING MEDIUM STORING MACHINE LEARNING PROGRAM, INFORMATION PROCESSING DEVICE, AND MACHINE LEARNING METHOD

Simplified Explanation

The patent application describes a machine learning program stored on a computer-readable medium that trains a machine learning model by combining multiple modules configured by a neural network. The process involves parsing word dependencies in question sentences related to images, determining weights for each module based on the parsing results, and selecting modules for the model based on these weights.

• Machine learning program for training a model with neural network modules:

 - The program combines modules created by a neural network to train a machine learning model.
 - It parses word dependencies in question sentences related to images to enhance model training.
 - Weights are assigned to modules based on parsing results to optimize their contribution to the model.

Potential Applications

This technology could be applied in: - Image recognition systems - Natural language processing tasks - Virtual assistants and chatbots

Problems Solved

This technology addresses issues such as: - Improving the accuracy of machine learning models - Enhancing the understanding of complex questions related to images - Optimizing the selection and utilization of neural network modules

Benefits

The benefits of this technology include: - Increased efficiency in training machine learning models - Enhanced performance in image and text-related tasks - Improved overall accuracy and reliability of AI systems

Potential Commercial Applications

Potential commercial applications of this technology include: - Developing advanced AI systems for various industries - Enhancing customer service through intelligent chatbots - Improving image recognition capabilities in security systems

Possible Prior Art

One possible prior art for this technology could be the use of neural networks in machine learning models for natural language processing tasks. Researchers have explored similar approaches in optimizing neural network modules for specific tasks, such as image recognition and text analysis.

How does this technology compare to existing methods in training machine learning models with neural network modules?

This technology introduces a novel approach by incorporating word dependency parsing in question sentences related to images. It would be interesting to see how this method compares to traditional training techniques in terms of model accuracy and efficiency.

What are the potential limitations or challenges in implementing this technology in real-world applications?

While the patent application outlines a promising method for training machine learning models, there may be challenges in scaling this approach to large datasets or complex tasks. Understanding the practical implications and limitations of this technology is crucial for its successful implementation.

Original Abstract Submitted

A non-transitory computer-readable recording medium storing a machine learning program for causing a computer to execute a process to train a machine learning model constructed by combining a plurality of modules configured by a neural network, the process includes parsing dependency between words for a plurality of words included in a question sentence of training data that forms a set of an image and the question sentence related to the image, determining a weight to be applied to each of the plurality of modules, based on a result of the parsing, and controlling selection of a combination of modules to be used in the machine learning model from the plurality of modules, based on the weight to be applied to each of the plurality of modules.