This assignment demonstrates how to use decision trees for classification and understand the foundational concepts behind ensemble methods such as random forests and boosted trees.
It is part of Week 4 (Course 2: Advanced Learning Algorithms) from the Machine Learning Specialization by Andrew Ng on Coursera.
In this lab, I implemented and trained a decision tree classifier using real-world data, explored how decision boundaries are formed, and visualized decision paths. I also learned how ensembles like bagging and boosting build on simple tree models to improve prediction accuracy and robustness.
- Decision tree construction and splitting criteria
- Tree depth, overfitting, and pruning
- Visualizing decision boundaries
- Introduction to tree ensemble methods (Random Forest, Gradient Boosting)
- Trade-offs between bias and variance in tree models
decision_tree_lab.ipynb: Main notebook with all exercises and visualizationslab_utils_tree.py: Helper functions for plotting, metrics, and decision boundariesdata/: Dataset files used for training and evaluation
⚠️ This repository includes only my own implementation and adheres to Coursera’s Honor Code.
- Python 3
- NumPy
- scikit-learn
- Matplotlib
- Jupyter Notebook
This assignment is part of:
Machine Learning Specialization
Instructor: Andrew Ng
Course 2: Advanced Learning Algorithms
Week 4: Tree Ensembles – Decision Trees, Random Forests, Boosting
This repository is intended for educational and portfolio use only. Not for direct submission on Coursera.
Star the repo if you're building tree-based models or exploring ensemble learning!