5-bostonhousing-mixturemodel-regression

简介

• 利用 Boston houseing 属性预测房价, 变量可能会存在交互作用
• 所以考虑使用混合模型

1. load package

Code

import MLJ:predict
using MLJ

2. load data

Code

X, y= @load_boston;

3. MLJ workflow

3.1 define model

Code

modelType= @load GaussianMixtureRegressor pkg = "BetaML"
gmr= modelType()

(fitResults, cache, report) = MLJ.fit(gmr, 1, X, y);

import BetaML ✔
Iter. 1:    Var. of the post  21.74887448784977       Log-likelihood -21687.09917379566

[ Info: For silent loading, specify `verbosity=0`. 

3.2 results

Code

y_res= predict(gmr, fitResults, X)
rmse(y_res,y)

7.9566567641159605