Mastering pandas - Second Edition

XGBoost case study

XGBoost is an ensemble algorithm popular for its outstanding performance. An ensemble algorithm involves multiple models instead of just one. Ensemble algorithms are of two types:

• Bagging: Here, a result from the algorithm is the average of results from individual models.
• Boosting: Here, we start from a base learner model. Each successive model gets created with better-trained parameters. The learning of new parameters happens through optimization algorithms such as gradient descent.

Next, we will look at the application of XGBoost on a dataset to predict the testing time of a newly manufactured car.

It is a step-by-step guide and you can just follow along:

1. Import the required packages:

# In[1]:    import os
    import pandas ...