What is XGBoost?

XGBoost: Unlocking Advanced Capabilities for Machine Learning Excellence

Javier Calderon Jr
4 min readOct 6, 2023



Certain tools stand out for their efficiency, versatility, and robustness. XGBoost, a gradient boosting framework, is one such tool that has garnered significant attention and praise from the data science community. Its adaptability to various tasks, from regression to classification, and its seamless integration with platforms like PySpark, make it an indispensable asset for any data scientist. In this article, we’ll delve deep into the capabilities of XGBoost and explore its prowess in handling multi-target regression, multi-label classification, multi-class tasks, batch training with external memory, and learning to rank.

Multi-Target Regression with XGBoost:

In real-world scenarios, predicting a single outcome often isn’t enough. For instance, predicting both the price and demand for a product simultaneously can provide more holistic insights. XGBoost’s multi-output regression capability allows for this, making it a go-to solution for such tasks.

import xgboost as xgb
# Prepare your data
data, targets = your_data_preparation_function()
# Create a multi-output regressor
model = xgb.XGBRegressor(objective ='reg:squarederror')
model.fit(data, targets)

Multi-Label Classification

Imagine classifying a news article into multiple categories like ‘Sports’, ‘International’, and ‘Entertainment’. Multi-label classification handles such scenarios, and XGBoost excels in this domain.

# Assuming binary relevance for simplicity
model = xgb.XGBClassifier(objective ='binary:logistic')
model.fit(data, label)

Multi-Class Classification

From image recognition to sentiment analysis, multi-class classification is omnipresent. XGBoost’s efficient handling of such tasks ensures accurate and fast results.

model = xgb.XGBClassifier(objective ='multi:softmax', num_class=n_classes)
model.fit(data, label)

Batch Training with External Memory



Javier Calderon Jr

CTO, Tech Entrepreneur, Mad Scientist, that has a passion to Innovate Solutions that specializes in Web3, Artificial Intelligence, and Cyber Security