XGBoost in ML Python - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding XGBoost Objective Functions

Which of the following objective functions is not commonly used in XGBoost for classification tasks?

A reg:squarederror

B multi:softmax

C binary:logistic

D count:poisson

💡 Hint

Think about which objective functions are for regression versus classification.

reg:squarederror is used for regression tasks, not classification. The others are classification or count models.

❓ Predict Output

intermediate

XGBoost Model Training Output

What will be the output of the following Python code snippet using XGBoost?

ML Python

import xgboost as xgb
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split

iris = load_iris()
X_train, X_test, y_train, y_test = train_test_split(iris.data, iris.target, test_size=0.2, random_state=42)
dtrain = xgb.DMatrix(X_train, label=y_train)
dtest = xgb.DMatrix(X_test, label=y_test)
params = {'objective': 'multi:softprob', 'num_class': 3, 'eval_metric': 'mlogloss'}
model = xgb.train(params, dtrain, num_boost_round=10, evals=[(dtest, 'eval')], verbose_eval=False)
preds = model.predict(dtest)
print(len(preds), len(preds[0]))

A 30

B 30 1

C 120 3

D 30 3

💡 Hint

Check the number of test samples and the shape of prediction output for multi-class softprob.

The test set has 30 samples. For 'multi:softprob', predictions are probabilities for each class, so output shape is (30, 3).

❓ Model Choice

advanced

Choosing XGBoost Parameters for Imbalanced Data

You have a highly imbalanced binary classification dataset. Which XGBoost parameter setting is best to help the model focus on the minority class?

A Set <code>max_depth</code> to a very high value like 20

B Set <code>scale_pos_weight</code> to the ratio of negative to positive samples

C Use <code>objective</code> as <code>reg:squarederror</code>

D Set <code>learning_rate</code> to 1.0

💡 Hint

Think about how to handle class imbalance in XGBoost.

scale_pos_weight helps balance the positive class weight, improving performance on imbalanced data. Other options are either irrelevant or harmful.

❓ Metrics

advanced

Evaluating XGBoost Model with Multi-class Log Loss

After training an XGBoost model with objective='multi:softprob', you get a multi-class log loss (mlogloss) of 0.8 on the test set. What does this value indicate?

A The model predictions have moderate confidence and some errors

B The model predictions are random and uninformative

C The model predictions are very confident and accurate

D The model predictions perfectly match the true labels

💡 Hint

Lower mlogloss is better; think about what 0.8 means.

A multi-class log loss of 0.8 means the model is somewhat confident but makes errors; perfect predictions would have mlogloss near 0.

🔧 Debug

expert

Debugging XGBoost Training Error

Consider this code snippet that raises an error during training:

import xgboost as xgb
import numpy as np

X = np.array([[1, 2], [3, 4], [5, 6]])
y = np.array([0, 1])
dtrain = xgb.DMatrix(X, label=y)
params = {'objective': 'binary:logistic'}
model = xgb.train(params, dtrain, num_boost_round=5)

What is the cause of the error?

A Missing eval_metric parameter in params

B Incorrect objective function for binary classification

C Mismatch between number of samples in X and y labels

D DMatrix requires labels to be float, not int

💡 Hint

Check the shapes of X and y carefully.

The feature matrix X has 3 samples but y has only 2 labels, causing a mismatch error during DMatrix creation.