國立虎尾科技大學 |

Machine Learning for Material Property Prediction and Analysis.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Machine Learning for Material Property Prediction and Analysis./
作者:	Magar, Rishikesh.
面頁冊數:	1 online resource (153 pages)
附註:	Source: Dissertations Abstracts International, Volume: 85-02, Section: B.
Contained By:	Dissertations Abstracts International85-02B.
標題:	Mechanical engineering. -
電子資源:	click for full text (PQDT)
ISBN:	9798380092876

Machine Learning for Material Property Prediction and Analysis.
Magar, Rishikesh.

Machine Learning for Material Property Prediction and Analysis. - 1 online resource (153 pages)

Source: Dissertations Abstracts International, Volume: 85-02, Section: B.

Thesis (Ph.D.)--Carnegie Mellon University, 2023.

Includes bibliographical references

The discovery of novel materials can significantly impact the quality of human life on the planet and make it more sustainable. Furthermore, novel materials can play a crucial role in technological advancements and enable innovations in industries such as healthcare and energy. However, the quest for new material discovery involves parsing the vast chemical space of materials. Navigating this vast chemical landscape to identify and synthesize materials with desirable properties is a big challenge. Considering the large scale of chemical space, developing advanced computational methods that can access this space effectively is very important. Computational methods that can accurately predict material properties can especially be leveraged for new material development. Accurate predictions can lead to the discovery of new materials with desirable properties for specific applications and reduce the time and cost associated with new material development. This dissertation investigates the potential of leveraging machine learning (ML) models as a more efficient computational tool for material property prediction and discovery.Traditionally, computational methods like Density Functional Theory (DFT) simulations have been used for high throughput screening and rapid prediction of material properties. Advancements in computational capabilities have enabled the large-scale usage of DFT simulations, yet parsing the entire chemical space and predicting the properties for all the possible materials is practically impossible due to the computational demands of these methods. Data-driven methods such as Machine Learning(ML) can learn patterns in the data and can be a viable alternative for DFT simulations. In general, ML models learn the material representations which can be leveraged to predict material properties at a fractional computational cost compared to DFT. Consequently, these models can be used as a screening tool to identify promising novel materials in the enormous chemical space which can then be further validated through experiments or DFT simulations.This dissertation explores various ML paradigms for material property prediction, encompassing Graph Neural Networks, Self-Supervised Representation Learning, Pretraining Strategies for Structure Agnostic Representation Learning, and Large Language Models. The primary aim of this work is to demonstrate the potential of ML models in accurate material property prediction and, as a result, catalyze novel material discovery through the application of machine learning.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798380092876Subjects--Topical Terms:

557493
Mechanical engineering.
Subjects--Index Terms:

Machine LearningIndex Terms--Genre/Form:

554714
Electronic books.

Machine Learning for Material Property Prediction and Analysis.
LDR:03997ntm a22004097 4500 001 1144276
005 20240531083816.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798380092876
035 $a (MiAaPQ)AAI30528636
035 $a AAI30528636
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Magar, Rishikesh. $3 1469252
245 1 0 $a Machine Learning for Material Property Prediction and Analysis.
264 0 $c 2023
300 $a 1 online resource (153 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-02, Section: B.
500 $a Advisor: Barati Farimani, Amir.
502 $a Thesis (Ph.D.)--Carnegie Mellon University, 2023.
504 $a Includes bibliographical references
520 $a The discovery of novel materials can significantly impact the quality of human life on the planet and make it more sustainable. Furthermore, novel materials can play a crucial role in technological advancements and enable innovations in industries such as healthcare and energy. However, the quest for new material discovery involves parsing the vast chemical space of materials. Navigating this vast chemical landscape to identify and synthesize materials with desirable properties is a big challenge. Considering the large scale of chemical space, developing advanced computational methods that can access this space effectively is very important. Computational methods that can accurately predict material properties can especially be leveraged for new material development. Accurate predictions can lead to the discovery of new materials with desirable properties for specific applications and reduce the time and cost associated with new material development. This dissertation investigates the potential of leveraging machine learning (ML) models as a more efficient computational tool for material property prediction and discovery.Traditionally, computational methods like Density Functional Theory (DFT) simulations have been used for high throughput screening and rapid prediction of material properties. Advancements in computational capabilities have enabled the large-scale usage of DFT simulations, yet parsing the entire chemical space and predicting the properties for all the possible materials is practically impossible due to the computational demands of these methods. Data-driven methods such as Machine Learning(ML) can learn patterns in the data and can be a viable alternative for DFT simulations. In general, ML models learn the material representations which can be leveraged to predict material properties at a fractional computational cost compared to DFT. Consequently, these models can be used as a screening tool to identify promising novel materials in the enormous chemical space which can then be further validated through experiments or DFT simulations.This dissertation explores various ML paradigms for material property prediction, encompassing Graph Neural Networks, Self-Supervised Representation Learning, Pretraining Strategies for Structure Agnostic Representation Learning, and Large Language Models. The primary aim of this work is to demonstrate the potential of ML models in accurate material property prediction and, as a result, catalyze novel material discovery through the application of machine learning.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Mechanical engineering. $3 557493
650 4 $a Materials science. $3 557839
653 $a Machine Learning
653 $a Material property prediction
653 $a Self-supervised learning
653 $a Density Functional Theory
653 $a Computational methods
653 $a Computational tool
655 7 $a Electronic books. $2 local $3 554714
690 $a 0800
690 $a 0794
690 $a 0548
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a Carnegie Mellon University. $b Mechanical Engineering. $3 1148639
773 0 $t Dissertations Abstracts International $g 85-02B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30528636 $z click for full text (PQDT)