Dataset Information

Machine learning approaches for predicting and validating mechanical properties of Mg rare earth alloys for light weight applications.

ABSTRACT: In this work, we have attempted to predict the mechanical behaviour of light weight Mg-based rare earth alloys fabricated through different mechanical and thermal processes. Our approach involves machine learning techniques across a range of different thermomechanical processes such as solution treatment, homogenization, extrusion and aging behaviour. The effectiveness of machine learning models is evaluated using performance metrics, including Coefficient of determination (R²), Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). After modeling and selection of best model, the mechanical behaviour of new alloys was predicted in terms of ultimate tensile strength, yield strength and total elongation. The predicted results highlight the superior predictive accuracy of the K-Nearest Neighbors (KNN) machine learning model, demonstrating its better performance metrics compared with other machine learning approaches. This model has been found to predict the material properties with an effective evaluation matrix (R² = 0.955, MAE = 3.4% and RMSE = 4.5%).

SUBMITTER: Jain S

PROVIDER: S-EPMC11792136 | biostudies-literature | 2025

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning approaches for predicting and validating mechanical properties of Mg rare earth alloys for light weight applications.

Jain Sandeep S Bhowmik Ayan A Lee Jaichan J

Science and technology of advanced materials 20250131 1

In this work, we have attempted to predict the mechanical behaviour of light weight Mg-based rare earth alloys fabricated through different mechanical and thermal processes. Our approach involves machine learning techniques across a range of different thermomechanical processes such as solution treatment, homogenization, extrusion and aging behaviour. The effectiveness of machine learning models is evaluated using performance metrics, including Coefficient of determination (R<sup>2</sup>), Mean ...[more]

PMID: 39906548

Similar Datasets

Project description:Purpose: The goals of this study is to compare the transcriptome (RNA-seq) modulations in Saccharomyces cerevisiae exposed to two rare earth elements. Lanthanum and ytterbium were used as representative of light and heavy rare earth elements, respectively. Methods: mRNA were sequenced from Saccharomyces cerevisiae exposed to two different rare earth elements. Lanthanum and ytterbium were used as representative of light and heavy rare earth elements respectively. The transcriptome of S. cerevisiae was analysed after being exposed for one hour to the EC10 (Effective concentration 10 %) and the EC50 (Effective concentration 50 %) of lanthanum (50 and 160 µM) and ytterbium (6 and 8 µM). The sequence reads were trimmed using Trimmomativ v0.36.1 and FastQC v0.67 used for quality check. Reads passing the quality check were mapped on the reference genome (S288C R64-2-1 of 2015-01-31 from https://www.yeastgenome.org/) of S. cerevisiae using TopHat v2.1.1. Reads that were mapped on the reference genome were quantified using HTSeq-count v0.6.1p1. Finally, differential gene expression analysis between treatments was carried out using DESeq2 v1.14.1. Differentially expressed genes between conditions were obtained and expressed as log2-fold change with adjusted p-values calculated via a Benjamini-Hochberg test. A cut-off adjusted p-value of < 0.01 was applied. Results: The transcription of genes related to several crucial pathways was modulated in response to both REEs, such as oxidative-reduction processes, DNA replication, and carbohydrate metabolism. REE-specific responses involving the cell wall and the pheromone signalling pathways were highlighted, while these were not reported for other metals. REE exposure also modified the expression and abundance of several ion transport systems, for which strong discrepancies were observed between the two contrasted REEs. Conclusions: Our results demonstrate the discrepancies in yeast response to different rare earth elements (light vs heavy rare earth elements). This results are valuable to prioritize key genes and proteins involved in REE detoxification mechanisms that would deserve further characterisation to better understand the REE toxicity on the environment and human health.

Dataset Information

Machine learning approaches for predicting and validating mechanical properties of Mg rare earth alloys for light weight applications.

Publications

Machine learning approaches for predicting and validating mechanical properties of Mg rare earth alloys for light weight applications.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets