1. Obtain the correlation coefficient between various characteristics of corporate income tax
1. Training points
(1) Master the correlation analysis methods in Python, and compare each analysis method.
(2) Understand and be able to use Python to realize the correlation analysis of the relevant characteristics of corporate income tax forecasting.
(3) Interpret the correlation analysis results.
2. Requirement description
Carry out correlation analysis on the original features that affect corporate income tax, and interpret the correlation between original features and the correlation between original features and target features.
3. Implementation ideas and steps
(1) Find the Pearson correlation coefficient between the original data features.
(2) Judging the correlation between each feature.
2. Select the key features of corporate income tax forecasting
1. Training points
(1) Understand the Lasso regression model, master its applicable scenarios, advantages and disadvantages.
(2) Master the method of feature selection using Lasso regression.
(3) Understand and master the Python code implementation of the above process.
2. Requirements Description
Filter the characteristics of factors that affect corporate income tax, and select the features that have a key impact on corporate income tax, laying the foundation for the next step of model construction.
3. Implementation ideas and steps
(1) Establish a Lasso regression model.
(2) Interpret the Lasso regression results.
3. Building a corporate income tax forecasting model
1. Training points
(1) Understand the gray forecasting model, and master its applicable scenarios, advantages and disadvantages.
(2) Understand the origin of support vector regression prediction and master its use.
(3) Understand and master the evaluation indicators of regression prediction models.
(4) Master all the Python code implementations in this process.
2. Demand statement
Based on the feature selection of the factors affecting corporate income tax, establish a single feature gray prediction model and a support vector regression prediction model, predict the corporate income tax in 2014 and 2015, and evaluate the model .
3. Implementation ideas and steps
(1) Use the gray forecasting model to predict the value of each feature in 2014 and 2015.
(2) Establish support vector regression prediction model.
(3) Evaluate the enterprise income tax prediction model established above.
