Machine Learning Based Power Estimation for CMOS VLSI Circuits

Nowadays, machine learning (ML) algorithms are receiving massive attention in most of the engineering application since it has capability complex systems modeling using historical data. Estimation power for CMOS VLSI circuit various attributes is proposed passive learning-based technique. The method uses supervised method, which provides a fast and accurate estimation without affecting accuracy system. Power random forest algorithm relatively new. Accurate circuits estimated by model optimized tuned multiobjective NSGA-II algorithm. It inferred from experimental results testing error varies 1.4% to 6.8% terms Mean Square Error 1.46e-06 when compared BPNN. Statistical like coefficient determination (R) Root (RMSE) done proven that Forest best choice with high 0.99938, low RMSE 0.000116.