Lane E. Schultz, Yiqi Wang, Ryan Jacobs, Dane Morgan
{"title":"使用核密度估计确定机器学习模型的领域:材料特性预测中的应用","authors":"Lane E. Schultz, Yiqi Wang, Ryan Jacobs, Dane Morgan","doi":"arxiv-2406.05143","DOIUrl":null,"url":null,"abstract":"Knowledge of the domain of applicability of a machine learning model is\nessential to ensuring accurate and reliable model predictions. In this work, we\ndevelop a new approach of assessing model domain and demonstrate that our\napproach provides accurate and meaningful designation of in-domain versus\nout-of-domain when applied across multiple model types and material property\ndata sets. Our approach assesses the distance between a test and training data\npoint in feature space by using kernel density estimation and shows that this\ndistance provides an effective tool for domain determination. We show that\nchemical groups considered unrelated based on established chemical knowledge\nexhibit significant dissimilarities by our measure. We also show that high\nmeasures of dissimilarity are associated with poor model performance (i.e.,\nhigh residual magnitudes) and poor estimates of model uncertainty (i.e.,\nunreliable uncertainty estimation). Automated tools are provided to enable\nresearchers to establish acceptable dissimilarity thresholds to identify\nwhether new predictions of their own machine learning models are in-domain\nversus out-of-domain.","PeriodicalId":501211,"journal":{"name":"arXiv - PHYS - Other Condensed Matter","volume":"63 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Determining Domain of Machine Learning Models using Kernel Density Estimates: Applications in Materials Property Prediction\",\"authors\":\"Lane E. Schultz, Yiqi Wang, Ryan Jacobs, Dane Morgan\",\"doi\":\"arxiv-2406.05143\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Knowledge of the domain of applicability of a machine learning model is\\nessential to ensuring accurate and reliable model predictions. In this work, we\\ndevelop a new approach of assessing model domain and demonstrate that our\\napproach provides accurate and meaningful designation of in-domain versus\\nout-of-domain when applied across multiple model types and material property\\ndata sets. Our approach assesses the distance between a test and training data\\npoint in feature space by using kernel density estimation and shows that this\\ndistance provides an effective tool for domain determination. We show that\\nchemical groups considered unrelated based on established chemical knowledge\\nexhibit significant dissimilarities by our measure. We also show that high\\nmeasures of dissimilarity are associated with poor model performance (i.e.,\\nhigh residual magnitudes) and poor estimates of model uncertainty (i.e.,\\nunreliable uncertainty estimation). Automated tools are provided to enable\\nresearchers to establish acceptable dissimilarity thresholds to identify\\nwhether new predictions of their own machine learning models are in-domain\\nversus out-of-domain.\",\"PeriodicalId\":501211,\"journal\":{\"name\":\"arXiv - PHYS - Other Condensed Matter\",\"volume\":\"63 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-05-28\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Other Condensed Matter\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2406.05143\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Other Condensed Matter","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2406.05143","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Determining Domain of Machine Learning Models using Kernel Density Estimates: Applications in Materials Property Prediction
Knowledge of the domain of applicability of a machine learning model is
essential to ensuring accurate and reliable model predictions. In this work, we
develop a new approach of assessing model domain and demonstrate that our
approach provides accurate and meaningful designation of in-domain versus
out-of-domain when applied across multiple model types and material property
data sets. Our approach assesses the distance between a test and training data
point in feature space by using kernel density estimation and shows that this
distance provides an effective tool for domain determination. We show that
chemical groups considered unrelated based on established chemical knowledge
exhibit significant dissimilarities by our measure. We also show that high
measures of dissimilarity are associated with poor model performance (i.e.,
high residual magnitudes) and poor estimates of model uncertainty (i.e.,
unreliable uncertainty estimation). Automated tools are provided to enable
researchers to establish acceptable dissimilarity thresholds to identify
whether new predictions of their own machine learning models are in-domain
versus out-of-domain.