Project 02 - Implementation of Machine Learning algorithms

This project consists of implementing solutions to different problems using machine learning algorithms (machine learning - ML). The project is divided into two themes:

• Supervised Learning: Three automatic classification methods (K-NN, Decision Tree and Random Forest) will be presented to solve a specific problem using the same database. The objective is to convey the usefulness of these methods for a more real problem and obtain important knowledge in this field.
• Unsupervised Learning: In this second part, two different problems will be presented on the same database, one of which will be solved using the clustering K-Means method and the other will be solved with association rules and the Apriori algorithm. The objective of this part is also to implement the use of these algorithms for more realistic situations and also to obtain some experience in the field.

This project is associated with the notebook ML_Notebook.pynb and to run it correctly, despite already being mentioned in it, it will be necessary to install the following packages using PIP:

1. pip install pandas
2. pip install numpy
3. pip install -U scikit-learn
4. pip install matplotlib
5. pip install mpl-toolkits.clifford
6. pip install seaborn
7. pip install mlxtend
8. pip install imblearn

To do this, you can use the command pip install -r requirements.txt in the appropriate environment to install all the packages mentioned above and run the Notebook related to Project 02.

In the end, the user will have learned the following knowledge:

• What is machine learning;
• What is a Dataset/Database;
• What is supervised learning;
• What is unsupervised learning;
• How to apply the two types of machine learning discussed;
• What is the k-nn method;
• What is the decision tree method;
• What is the random forest method;
• What is clustering and the k-means method;
• What are association rules and the apriori method;

Running and reading each notebook should be more than enough to solidly introduce the user to the basic concepts of Artificial Intelligence. Solving more real problems makes the association of concepts easier and will allow future projects to be based on the themes in the same structure.

---

This work was authored by:

• David Rodrigues ✅
• André Araújo ✅

Supervised by:

• Professor Joaquim Silva ✅

If you have any doubts or questions:

• Leave a question on the issues tab
• E-mail us at a21177@alunos.ipca.pt (David) and a21185@alunos.ipca.pt (André)