English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This study focuses on developing precise machine learning (ML) regressionmodels for predicting energy bandgap values based on chemical compositions andcrystal structures. The primary aim is to match the accuracy of predictionsderived from GGA-PBE calculations and validate them through density functionaltheory (DFT)-based band structure calculations. We assessed eight standalone MLregression models, including AdaBoost, Bagging, CatBoost, LGBM, RF, DT, GB, andXGB. These models were analyzed for their ability to predict GGA-PBE bandgapvalues across diverse material structures and compositions, using a datasetcontaining bandgap values for 106,113 compounds. Additionally, we constructedfour ensemble models using the stacking method and seven using the baggingmethod. These ensemble models incorporated RidgeCV and LassoCV to explore ifensemble techniques could enhance prediction accuracy. The dataset was dividedinto subsets of varying sizes: 10,000, 25,000, 50,000, and 100,000 entries. Wedetermined feature importance through permutation techniques and established acorrelation coefficient matrix using the Pearson correlation method. The RandomForest (RF) model emerged as the top performer among standalone models,achieving an R2 value of 0.943 and an RMSE value of 0.504 eV. Baggingregression demonstrated improved performance across different dataset sizeswith streamlined feature selection. Ensemble models, particularly bagging,consistently outperformed standalone models, achieving the best R2 value of0.948 and an RMSE value of 0.479 eV in the test dataset. Using thebest-performing model, we predicted bandgap values for new half-Heuslercompounds with 18 valence electron counts. These predictions were successfullyvalidated using accurate DFT calculations. DFT calculations indicated that thenewly predicted compounds are narrow bandgap semiconductors with dynamicstability.

Composition and Structure Based GGA Bandgap Prediction Using Machine Learning Approach