Overview
Explore techniques to identify and prevent overfitting in neural networks in this 26-minute video tutorial. Learn about early stopping, audio data augmentation, dropout, and L1/L2 regularisation. Follow along as the instructor implements dropout and regularisation in a music genre classifier. Access accompanying slides and code on GitHub for hands-on practice. Gain insights into simple architecture, audio data documentation, dropout probability, and regularization examples. By the end of the tutorial, acquire practical skills to improve neural network performance and prevent overfitting in audio-based machine learning projects.
Syllabus
Introduction
Identifying Overfitting
Results
Overview
Simple Architecture
Audio Data Documentation
Early Stop
Dropout
Dropout probability
Regularization
Regularization Examples
Coding
Conclusion
Taught by
Valerio Velardo - The Sound of AI