Data Mining for Chinese Left-behind Children Problem
The project implement feature selection and modelling for left-behind children problem by Python. Feature selection consists of filter and wrapper approaches. Models contains support vector regression(SVR) and neural network regression(NNR).
Filter approach is included in 3.filter method for features selection.
The models built using filtered features are in 4.svr/1.model by filter method, 5.nnr/1.model by filter method.
Wrapper approach and corresponding models are included in 4.svr/2.genetic algorithm, 4.svr/3.clustering method, 5.nnr/2.genetic algorithm and 5.nnr/3.clustering method.
7.models contains several selected SVR models.