可靠和可解释的机器学习方法加速材料发现
Reliable and Explainable Machine Learning Methods for Accelerated Material Discovery
摘要:
材料科学家越来越多地采用机器学习(ML)来做出潜在的重要决策,例如材料的发现、开发、优化、合成和表征。然而,尽管ML在商业应用中的表现令人印象深刻,但在材料科学应用中应用ML时存在一些独特的挑战。在这种情况下,这项工作的贡献是双重的。首先,我们确定了现有ML技术在从代表性不足/不平衡的材料数据中学习时的常见陷阱。具体来说,我们表明,在不平衡数据的情况下,评估ML模型质量的标准方法会失效,并导致误导性的结论。此外,我们发现模型自己的置信度得分是不可信的,模型内省方法(使用更简单的模型)也没有帮助,因为它们会导致预测性能的损失(可靠性-可解释性权衡)。其次,为了克服这些挑战,我们提出了一个通用的可解释和可靠的机器学习框架。具体来说,我们提出了一种新的管道,采用一个更简单的模型,可靠地预测材料性能的合奏。我们还提出了一种迁移学习技术,并表明,由于模型的简单性的性能损失可以克服利用不同的材料属性之间的相关性。一个新的评估指标和信任分数,以更好地量化的预测的信心也提出了。为了提高可解释性,我们添加了一个合理的发电机组件,我们的框架,提供模型级和决策级的解释。最后,我们证明了我们的技术在两个应用中的多功能性:1)预测晶体化合物的性质,2)确定新的潜在稳定的太阳能电池材料。
英文摘要:
Material scientists are increasingly adopting the use of machine learning (ML) for making potentially important decisions, such as, discovery, development, optimization, synthesis and characterization of materials. However, despite ML's impressive performance in commercial applications, several unique challenges exist when applying ML in materials science applications. In such a context, the contributions of this work are twofold. First, we identify common pitfalls of existing ML techniques when learning from underrepresented/imbalanced material data. Specifically, we show that with imbalanced data, standard methods for assessing quality of ML models break down and lead to misleading conclusions. Furthermore, we found that the model's own confidence score cannot be trusted and model introspection methods (using simpler models) do not help as they result in loss of predictive performance (reliability-explainability trade-off). Second, to overcome these challenges, we propose a general-purpose explainable and reliable machine-learning framework. Specifically, we propose a novel pipeline that employs an ensemble of simpler models to reliably predict material properties. We also propose a transfer learning technique and show that the performance loss due to models' simplicity can be overcome by exploiting correlations among different material properties. A new evaluation metric and a trust score to better quantify the confidence in the predictions are also proposed. To improve the interpretability, we add a rationale generator component to our framework which provides both model-level and decision-level explanations. Finally, we demonstrate the versatility of our technique on two applications: 1) predicting properties of crystalline compounds, and 2) identifying novel potentially stable solar cell materials.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。