ModelAuto

Model Automation

This library will help in automating the processes of Model making (Regression / Classification) ,In few lines of code. One of the major parts are Selection of Importent Features and Model . All the Modules are properly documented ,You can find it here

The modules help in -

1. Data Preprocessing
2. Removing Multicolinierity
3. Selecting Features 😀
4. Selecting best Model 😉 (Reg- LR , SVM ,… Classif- SVM ,LogR , …)

Most of the modules have graphical techniques for better decision making at Selecting Features and Models

Link to Video Demo

Link to PyPi

Description

• Usage

  • Getting Started
  • Required Libraries
  • Demo Code - (Kaggle Housing Price)

• Demo on Kaggle - Housing Prices (Advance Regression)

• Conclusion

Usages

• Getting Started

  
       ' Model Automation '
  
        pip install ModelAuto
  
  

  
  

• Required Libraries

  1. Sklearn
  2. seaborn
  3. statsmodels

• Demo Code for complete model

  
        from ModelAuto.Datapreprocess import Preprocessing
        from ModelAuto.FeatureSelection import backwardElimination
        from ModelAuto.ModelSelection import Regress_model
  
        DATA = pd.read_csv('Path/to/file.csv')
          
  
                            ''' 1. Data Preprocessing '''
  
        X_train , X_test = Preprocessing( X_train, X_test , Multi =True)
  
                            ''' 2. Select best Features '''
  
        Features = backwardElimination( x_train , y_train )
          
                            ''' 3.  Select best Model    '''
  
        x_train, x_test, y_train, y_test = train_test_split( Features , y_train, test_size=0.2, random_state=1 )
  
        Model = Regress_model( x_train,x_test,y_train,y_test )
  
                            ''' 4. Make Predicatoins '''
  
        Predictions = Model.predict('Test Data')
  
          
                                print('Done !!!😀😀')
  
                                                  
                '''For more flexibility in Data Preprocessing  '''
  
  
        x_data = handel_nan(x_data)
  
        y_data = handel_nan(y_data)
  
        x_train, x_test = handel_Standerd( x_train , x_test )
  
        x_train , x_test = handel_Catagorical( x_train , x_test )
  
  
  

Demo on KAGGLE - Housing Prices (Advance Regression)

Link to Kaggle Competion

A detailed documentation is given on the various modules. In this documentation implimentation on the Kaggle Competition have been shown. We will be looking at-

How to use various Feature selection methods and Model selection methods.

Link to the Documentation

Conclusion

Basically it

1. Speeds up the Model making process .

2. It gives us a clear understanding of the complete model which helps in looking at the minute changes which can boost the model efficiency.

3. A person with very basic knowledge could also create efficient models.

4. The Vizvalization techniques help in understanding concepts easily.

This site is open source. Improve this page.