View Categories

13. Statistical Analysis of Training Data

Once the user uploads the training data in a new analysis task, statistical analysis can be performed prior to model training. In the “Transform Data” section, fundamental data analysis and correlation statistics of the training data can be accessed by clicking the “Generate Profiling” button. This file contains histograms of the data distribution for each column and the correlation coefficients between the data in each column.

Below is an example of a profiling HTML file:

*This functionality is implemented using the features of pandas-profiling.

https://github.com/ydataai/pandas-profiling