XGBoost: Intro, Step-by-Step Implementation, and Performance Comparison
XGBoost has become one of the most popular well-rounded regressors and/or classifiers for all machine learning practitioners. If you ask a data scientist what model they would use for an unknown task, without any other information, odds are they will choose XGBoost given the vast types of use cases it can be applied to — it is quick, reliable, versatile and very easy to implement.
Today, we will conceptually review XGBoost, walk through the implementation of it and compare its performance to several other models for a classification task.
1. XGBoost — Conceptual Overview
XGBoost stands for Extreme Gradient Boosting. It is a gradient boosting decision tree type of a model, that can be used both for supervised regression and classification tasks. We used a few terms to define XGBoost so let’s walk through them one by one to better understand them. I will provide a brief overview of each concept and will include links to other posts with more details for those interested in a deeper discussion.
Supervised vs. Unsupervised Learning: Supervised is when we use labled data to train a machine learning (ML) model. Therefore, the model can learn from the labels of the data that we provided. On the other hand, unsupervised learning is when our data does not have lables and the model learns the underlying patterns in the provided data, in the absence of any labels.
0 Comments