Find the people can help me in build a model (symbolic regression) + giải thuật di truyền (Genetic Programming) Know how to use (DEAP) trên python
Budget: $30 – $250 USD
i will describe my problem.
first we will ten function f1 = b1x1+....b8x8 , f10 = b1x1+....b8x8 with b is the coefficents and x is (factors effect to f the compressive strength concrete). We will build a model by machine learning with the data 1030 rows and get the final y = b1f1+ b2f2 + .... b10f10. But it not good right now and we need to optimize it. we will use DEAP (open source library in python to help run genetic programming in treebase structure) to generate many offspring and evaluate all of offspring and ranking it we will pick ten best function in offspring and input it with ten last function above (total 20 functions) build model again and get the other y = b1f1 +......+ b20f20. after that we evaluate contribution of (b) and ranking it. Finally, just 10 best functions were chosen and build the model finally. if RMSE of finally model < RMSE (ten initial functions) we will replace the best function and the best RMSE, if not we keep RMSE and 10 initial function and start the other loops.
first we will ten function f1 = b1x1+....b8x8 , f10 = b1x1+....b8x8 with b is the coefficents and x is (factors effect to f the compressive strength concrete). We will build a model by machine learning with the data 1030 rows and get the final y = b1f1+ b2f2 + .... b10f10. But it not good right now and we need to optimize it. we will use DEAP (open source library in python to help run genetic programming in treebase structure) to generate many offspring and evaluate all of offspring and ranking it we will pick ten best function in offspring and input it with ten last function above (total 20 functions) build model again and get the other y = b1f1 +......+ b20f20. after that we evaluate contribution of (b) and ranking it. Finally, just 10 best functions were chosen and build the model finally. if RMSE of finally model < RMSE (ten initial functions) we will replace the best function and the best RMSE, if not we keep RMSE and 10 initial function and start the other loops.