Laboratory synthesis results prediction
The synthesis product must have specific properties, such as viscosity and other characteristics. It was necessary to achieve optimal parameters of its properties on the existing production line and, within the framework of a given infrastructure, select the optimal properties of the technical process (select a different mixture of initial components, a different temperature regime, or change other parameters). Three regressors were built. Then, with the help of these regressors, they were presented in the form of a function, and with the help of genetic algorithms, the optimal parameters were found according to the prediction results of the regressor.
python; numpy, pandas, matplotlib, seaborn, nltk, spacy, scikit-learn, keras; Git, GitHub