CONNECT WITH US:
Sep 30, 2021 07:00 JST
Science and Technology of Advanced Materials
Improving machine learning for materials design
A quick, cost-effective approach improves the accuracy with which machine learning models can predict the properties of new materials.
TSUKUBA, Japan, Sep 30, 2021 - (ACN Newswire) - A new approach can train a machine learning model to predict the properties of a material using only data obtained through simple measurements, saving time and money compared with those currently used. It was designed by researchers at Japan's National Institute for Materials Science (NIMS), Asahi KASEI Corporation, Mitsubishi Chemical Corporation, Mitsui Chemicals, and Sumitomo Chemical Co and reported in the journal Science and Technology of Advanced Materials: Methods.
The new approach can predict difficult-to-measure experimental data such as tensile modulus using easy-to-measure experimental data like X-ray diffraction. It further helps design new materials or repurpose already known ones.
"Machine learning is a powerful tool for predicting the composition of elements and process needed to fabricate a material with specific properties," explains Ryo Tamura, a senior researcher at NIMS who specializes in the field of materials informatics.
A tremendous amount of data is usually needed to train machine learning models for this purpose. Two kinds of data are used. Controllable descriptors are data that can be chosen without making a material, such as the chemical elements and processes used to synthesize it. But uncontrollable descriptors, like X-ray diffraction data, can only be obtained by making the material and conducting experiments on it.
"We developed an effective experimental design method to more accurately predict material properties using descriptors that cannot be controlled," says Tamura.
The approach involves the examination of a dataset of controllable descriptors to choose the best material with the target properties to use for improving the model's accuracy. In this case, the scientists interrogated a database of 75 types of polypropylenes to select a candidate with specific mechanical properties.
They then selected the material and extracted some of its uncontrollable descriptors, for example, its X-ray diffraction data and mechanical properties.
This data was added to the present dataset to better train a machine learning model employing special algorithms to predict a material's properties using only uncontrollable descriptors.
"Our experimental design can be used to predict difficult-to-measure experimental data using easy-to-measure data, accelerating our ability to design new materials or to repurpose already known ones, while reducing the costs," says Tamura. The prediction method can also help improve understanding of how a material's structure affects specific properties.
The team is currently working on further optimizing their approach in collaboration with chemical manufacturers in Japan.
National Institute for Materials Science (NIMS)
About Science and Technology of Advanced Materials: Methods (STAM Methods)
STAM Methods is an open access sister journal of Science and Technology of Advanced Materials (STAM), and focuses on emergent methods and tools for improving and/or accelerating materials developments, such as methodology, apparatus, instrumentation, modeling, high-through put data collection, materials/process informatics, databases, and programming.
Dr. Yoshikazu Shinohara
STAM Methods Publishing Director
Press release distributed by Asia Research News for Science and Technology of Advanced Materials.
Source: Science and Technology of Advanced Materials
Sectors: Science & Nanotech
Copyright ©2024 ACN Newswire. All rights reserved. A division of Asia Corporate News Network.
Related Press Release
Sensing structure without touching
February 27 2024 08:00 JST
Nano-sized probes reveal how cellular structure responds to pressure
November 21 2023 07:00 JST
Machine learning techniques improve X-ray materials analysis
November 17 2023 10:00 JST
A bio-inspired twist on robotic handling
November 14 2023 20:00 JST
GPT-4 artificial intelligence shows some competence in chemistry
October 17 2023 08:00 JST
Closing the loop between artificial intelligence and robotic experiments
August 24 2023 09:00 JST
Machine intelligence for designing molecules and reaction pathways
May 24 2023 09:00 JST
Face-down: Gravity's effects on cell movement
May 13 2023 00:00 JST
Polymer protection for vaccines and drugs
December 09 2022 23:00 JST
Revealing crystal structures robotically
December 02 2022 20:00 JST
More Press release >>
Latest Press Release
DOCOMO to Conduct Open RAN Field Trials with Ooredoo, Smart and StarHub
Feb 27, 2024 19:53 JST
CleverTap partners with Zoomcar to drive customer engagement on their app
Feb 27, 2024 18:00 JST
NEC Selected by NTT DOCOMO as a Virtualized Radio Access Network (vRAN) Vendor
Feb 26, 2024 17:54 JST
Fujitsu and Celonis Expand Strategic Global Partnership
Feb 26, 2024 15:40 JST
Mitsubishi Corporation's Sale of Exportadora de Sal, S.A. de C.V., a Salt Production Company in Mexico
Feb 26, 2024 14:29 JST
'MMVO,' Mazda's Production Base in Mexico Marks its 10th Anniversary
Feb 26, 2024 14:00 JST
Mitsubishi Electric Building Solutions to Supply Elevators and Escalators for New Capital Relocation Project in Indonesia
Feb 26, 2024 11:00 JST
Clear Mobitel Successfully Deploys NEC 5G SA Cloud Native Core Network to Accelerate the Adoption of Innovative 5G Solutions in the UK and the Channel Islands
Feb 22, 2024 17:13 JST
NEC UPF Achieves Industry-Leading High Performance in 5G Core Networks, reaching 1.3Tbps Throughput
Feb 22, 2024 16:10 JST
DOCOMO and NTT Expand 6G Collaborations with SK Telecom and Rohde & Schwarz
Feb 22, 2024 15:33 JST
Successful demonstration of computing and mobile networks convergence to provide diverse services in the 6G era
Feb 22, 2024 13:18 JST
Fujitsu introduces "Uvance Wayfinders", expanded and strengthened consulting capabilities to deliver cross-industry business value
Feb 22, 2024 12:05 JST
Fujitsu delivers new supercomputer system to Japan Meteorological Agency to improve prediction accuracy for typhoons and torrential rain
Feb 21, 2024 16:38 JST
MHI Group Presents "Best Innovation 2023" Awards for Activities that Contribute to Protecting the Environment
Feb 21, 2024 16:28 JST
Eisai to Boost Initiatives on Greenhouse Gas Reduction, Aiming to Achieve Net Zero by 2050
Feb 21, 2024 15:26 JST
NEC, Arm, Qualcomm, Red Hat, and Hewlett Packard Enterprise demonstrate Open vRAN and 5G Core UPF using Arm-based CPUs
Feb 21, 2024 14:09 JST
NEC launches new User Plane Function (UPF) product for telecom operators, aiming to achieve networks for the Beyond 5G/6G era
Feb 21, 2024 11:33 JST
TANAKA to Provide Pure Gold, Pure Silver, and Pure Bronze Medals for the Tokyo Marathon 2024
Feb 21, 2024 04:00 JST
Fujitsu collaborates with QuTech in development of new technology for freezing electronics to control diamond spin qubits
Feb 20, 2024 21:28 JST
NTT DOCOMO and SK Telecom Release White Paper on Base Station Equipment Utilizing Virtualization Technology
Feb 20, 2024 15:32 JST
More Latest Release >>