Black liquor valorization Machine learning model
Description
This data set used microbial fermentation system adopting constant initial COD or constant sucrose COD with BL COD supplementation. Each batch have multiple sets with replicate. This data set is used to model COD consumption, biomass, polymer production. The model is further used to optimize the yield.
Files
Steps to reproduce
a0_all_analysis is the main function referring all other functions in the folder. 'a1_dataload.m' is used to load the experimental data in excel files, store it as "a1_Summary_Polymer". The a2_megaplot is used to have stacked error bar plot, while a3_myheatanalysis maps correlation heatmap. The a4_ANN_Bayesopt.mat and a4_RF_Bayesopt is used to build the the ANN and RF model with hyperparameter optimization. a6_modelscatter is used to generate scatter plot between experimental and modeled responses. a8_LIME_Shaply_finalplot.mat is LIME and Shapley analysis. b1_min_max_GA used to get individual optimization of responses. b3_ANN_multiobj is used for multi objective optimization
Institutions
- Karunya UniversityTamil Nadu, Coimbatore