Skip to content

IAT-ExploringAI-2024/Week4-NeuralNetworks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 

Repository files navigation

Neural Networks

Neural networks are computational models inspired by the human brain. They consist of layers of interconnected nodes, or "neurons," each layer designed to perform specific types of transformations on its input data. Typically, a neural network includes:

Input Layer: This is the first point of data entry, where each node represents one feature of the input data.

Hidden Layers: These layers perform complex transformations on the input data using weights and activation functions. Deeper networks with more hidden layers can learn more complex patterns.

Output Layer: The final layer produces the network's output, such as a classification or prediction. NEURAL NETWORK

The power of neural networks lies in their ability to learn the appropriate transformation weights through a process called "training," often using backpropagation and gradient descent algorithms. This enables them to make predictions or recognize patterns in unseen data, making them highly effective for tasks like image and speech recognition, and in your case, audio data classification.

Audio Data Classification Assignment USING NN

Overview

This assignment's focus is on implementing neural networks to model audio data that you created previously. This follows our previous exploration of classical machine learning models using Random Forest, SVM and KNN.

Assignment Tasks

Tasks

  1. Fork the repository to your own account.
  2. Perform audio data classification utilizing neural network model on RAVDESS.
  3. Evaluate the model both in the original test-set and also using your new audio data. Create and compare the confsion matrix for both the original test data, and your audio data.
  4. Commit your changes to your repository.
  5. Create a detailed 1–2-page report (500-1000 words) on the modeling you've undertaken and your results. Compare the accuracy on your own test set and RAVDESS test set, use graphics of the confusion matrix and results of different metrics. Compare differences in performance, accuracy, and computational efficiency between neural networks and classical approaches you had from last week, and provide insights into why these differences occur.
  6. In your report, illustrate the comparsions via graphs.
  7. Upload your report on Canvas in PDF format, along with the link of your repository and data. This will provide a comprehensive overview of your exploration and application of various techniques in audio data classification modeling.

Contact

For queries or further discussions, feel free to reach out to TA.

Credits: https://github.com/IliaZenkov/sklearn-audio-classification

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published