ML model prediction to optimize composition of cellulose based film

This data set contains properties of cellulose-based film prepared with different proportions of PEG, malic acid, and hexdecanoate. Four attributes of the film are modeled with Fitlm (ANN) and ensemble modeling tools (RF) while optimizing hyperparameters. The composite desirability-based optimization workflow is presented.本数据集包含采用不同比例聚乙二醇（PEG）、苹果酸（malic acid）与十六烷酸酯（hexdecanoate）制备的纤维素基薄膜的性能参数。针对该薄膜的四项属性，本研究采用Fitlm（人工神经网络ANN）与集成建模工具随机森林（RF）开展建模，并同步优化超参数。本文同时提出了基于复合期望函数的优化工作流程。