I trained my Decision tree classifier model with the Titanic dataset. I tested the model it is fine only. All went well. But then I learned something called visualization of training (learning curve and validation curve) which will be useful to analyze how my model gets trained. For TensorFlow or Keras model, there is a visualization tool called tensorboard which is great but there is no provision for sklearn model. I searched on the internet but got something called sklearn.model_selection.learning_curve and sklearn.model_selection.validation_curve but I tried didn't get satisfied. Then I searched in internet and got to know about Weight & bias which is a very wonderful and great tool for visualization. They support sklearn, PyTorch, TensorFlow, etc. In this blog, I will tell how to integrate it in your project.
Installation
To install the weight and bias add the following line
To install using python
pip install wandb
To install using anaconda
conda install wandb
Set up
Import W&B and Login
import wandb
wandb.login()
In this article, I am going to show how to integrate W&B for all regression, classification and clustering and here we use ridge model for regression and classification and kmeans for clustering