UniPD Machine Learning Course Exercises

This repository contains materials and code implementations for the Machine Learning labs conducted during the academic year 2022-23.

Linear Classification and Regression

Tasks:

• Classification: Implement the Perceptron algorithm and use Logistic Regression from Python libraries for NBA players' roles classification.
• Regression: Use Least Squares implementation in Python libraries for house prices data regression.

Classification with SVM

Task:

• Classify ancient cursive Japanese (Kuzushiji) writing using Support Vector Machines (SVM).

Instructions:

• Classify images of characters using SVM with different kernels.
• Estimate parameters with cross-validation and visualize results with confusion matrices.

Classification with Neural Networks

Task:

• Classify ancient cursive Japanese (Kuzushiji) writing using Neural Networks (NN).

Deep Learning with Keras Tutorial

Welcome to the Deep Learning with Keras Tutorial repository! This repository contains materials and code examples from our tutorial on implementing deep neural networks using Keras. Whether you're a beginner or an experienced practitioner, these resources will help you dive into the world of deep learning.

Overview

In this tutorial, we cover the fundamentals of deep learning using Keras, a high-level neural networks API, written in Python and capable of running on top of TensorFlow. We explore various topics, including:

• Introduction to Keras and Deep Learning: Understand the basics of Keras and its advantages in building deep neural networks.

• Sequential vs. Functional Models: Learn the difference between sequential and functional models and when to use each approach.

• Building Convolutional Neural Networks (CNNs): Explore the construction of CNNs for image recognition tasks.

• Optimizers, Loss Functions, and Metrics: Understand the significance of optimizers, loss functions, and metrics in neural network training.

Contents

• Code Examples: Find Python scripts and Jupyter notebooks demonstrating different aspects of deep learning with Keras.

• Resources: Links to useful articles, tutorials, and documentation to deepen your understanding of Keras and neural networks.

Instructions:

• Use Neural Networks with different architectures and experiment with batch size and learning rate.
• Plot the estimated weights for analysis.

Prerequisites:

• Python 3.x
• Libraries: NumPy, SciPy, scikit-learn, matplotlib, Jupyter notebook/lab

Contributing

Contributions are welcome! If you find a bug or want to add new features or examples, please open an issue or submit a pull request Happy coding and happy learning! 🚀

License

This project is licensed under the MIT License - see the LICENSE.md file for details.