You can use AI-based regression models for example support vector regression or deep learning regression for parameter tuning. For example, consider a simple dc-dc boost converter. You can collect dataset on a dc-dc boost converter. Or in case you do not have a dataset, you can simulate a dc-dc boot converter for specific power rating and generate your own dataset. Then you can feed the dataset to a AI regression model to get design parameters, for example : inductor, and capacitor values. Then apply those design parameters to check for the efficacy of your design. In addition, you can compare the design parameter you got from AI with your own (manually calculated) design parameters to make a comparative analysis.