摘要
由一名新闻记者-机器人与机器学习的工作人员新闻编辑每日新闻-调查人员发布了关于马学习的新报告。根据中国人民代表大会北京的新闻报道，NewsRx记者的研究表明：“非线性光学(NLO)材料由于其固有的波长转换能力，在现代光学和工业中具有重要意义，带隙是NL O晶体的一个关键特性。”本研究的资助机构包括国家自然科学基金(NSFC)、国家自然科学基金(NSFC)、浙江省自然科学基金、北京师范大学创业基金。记者从北京师范大学的一篇研究中得到一句话:“近年来,机器学习(ML)已成为在合成前预测化合物带隙的有力工具,但现有的NLO晶体实验数据的不足,给利用ML技术开发新型NLO材料带来了巨大挑战。”本文提出了一种新的基于多级描述符(Z-Y.Zhan G,X.Liu,L.Shen,L.Chen and W.-H.Fang,J.Phys.Chem.C,2021,125,25175-2.5188)和梯度增强回归树算法的多质量ML方法,并将计算和实验的NLO晶体带隙作为低保真度和高保真度的标记。多保真度ML模型克服了单保真度预测器的性能，进一步观察到对低保真度La Bel的预测精度较低可能导致对高保真度标记的预测精度较高。用性能最好的多保真度ML模型在实验带隙测试集上的平均绝对误差为0.293 eV，比用单保真度模型(0.355 eV)小，这远远不够完善，但也不够准确，不足以作为发现新型非线性光学材料的第一步OL的有效计算方法。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on Ma chine Learning. According to news reporting originating in Beijing, People's Rep ublic of China, by NewsRx journalists, research stated, "Nonlinear optical (NLO) materials are of great importance in modern optics and industry because of thei r intrinsic capability of wavelength conversion. Bandgap is a key property of NL O crystals." Financial supporters for this research include National Natural Science Foundati on of China (NSFC), National Natural Science Foundation of China (NSFC), Natural Science Foundation of Zhejiang Province, Beijing Normal University Startup. The news reporters obtained a quote from the research from Beijing Normal Univer sity, "In recent years, machine learning (ML) has become a powerful tool to pred ict the bandgaps of compounds before synthesis. However, the shortage of availab le experimental data of NLO crystals poses a significant challenge for the explo ration of new NLO materials using ML. In this work, we proposed a new multi-fide lity ML approach based on the multilevel descriptors developed by us (Z.-Y. Zhan g, X. Liu, L. Shen, L. Chen and W.-H. Fang, J. Phys. Chem. C, 2021, 125, 25175-2 5188) and the gradient boosting regression tree algorithm. The calculated and ex perimental bandgaps of NLO crystals were collected as the low- and high-fidelity labels, respectively. The experimental values were predicted based on chemical compositions of crystals without prior knowledge about crystal structures. The m ulti-fidelity ML model overcame the performance of single-fidelity predictor. Fu rthermore, it was observed that less accurate predictions on the low-fidelity la bel may result in more accurate prediction on the high-fidelity label, at least in the present case. Using the multi-fidelity ML model with the best performance in this work, the mean absolute error on the test set of experimental bandgaps was 0.293 eV, which is smaller than that using the single-fidelity model (0.355 eV). It is far from perfect but accurate enough as an effective computational to ol in the first step to discover novel NLO materials."
NETL
NSTL