Scipy优化机器学习模型

Question

最近，我一直面临一个问题，我认为Scipy可能是一个很好的解决方案。但是，我无法正确应用它。不确定我是否缺少某些东西，或者我正在寻找的东西实际上根本不可能。

这是一个虚构的示例，我为使事情变得更加清晰，更易于可视化。我的情况更加复杂。但是，我想发现的是，会有多少狗和猫给出特定数量的老鼠？

from sklearn.svm import SVR
from scipy.optimize import minimize
from sklearn.ensemble import RandomForestRegressor
import numpy as np
import pandas as pd

n_dogs = [10, 5, 5, 2, 19, 12, 1,2]
n_cats = [5, 100, 5, 3, 1000, 0, 1,2]
n_rats = [100, 0, 50, 30, 0, 1000, 10, 5]


X = np.array([n_dogs, n_cats]).T
y = np.array([n_rats]).T 
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)    
model = RandomForestRegressor()
model.fit(X,y)

#Boundaries of condition
bnds = [(100,500), (100,150)]
#Initial guesses
x0 = [300,125]
#Predict from condition
residual_plant = minimize(lambda x: model.predict([[12,8]]), x0, method='SLSQP',bounds=bnds,options = {'eps': np.radians(5.0)})
print(residual_plant)```