Neural Web Tool

Provides an accessible and visually appealing learning experience for casual and serious users who wish to learn about neural networks.

Table of Contents:

1. Description

Description:

With the growing amount of interest and market presence regarding Artificial Intelligence, education on topics such as Deep Learning and Machine Learning have high market value. Making this material approachable is important, and online interactive tools such as our product are an excellent candidate for accomplishing this. Whether a student in a machine learning course, or an individual simply interested in simulating and learning about artificial neural networks, our product aims to make an accessible and rewarding experience. As an educator, our product can be used as an effective tool for assigning, managing, and grading homework assignments for students. We also recognize that our tool can be a helpful reference for hobbyists, professionals, and others who aim to expand their knowledge in manageable, accessible steps.

The application has two core components:

1. Neural Network Sandboxes

   In the sandbox environment, a user can create a networks from scratch, or use examples. By having full control over an artificial neural network, we effectively create a sandbox for users to learn about a rigorous topic in whatever way suits them best. Hovering over different components provide basic explanations of what they’re looking at, and each includes expandable reference material. Additionally there will be live data visualizations on the amount of iterations, system resource usage, and time to learn that will help users better correlate static parts of the neural network (neurons and layers) with dynamic, real-time information.

2. Learning Modules

   With learning modules, the user will follow a progression-based learning experience. The user will start with the basics of neurons and training methods, and then move up to full scale neural networks and data sets. As they progress, users will be given an interactable interface on which they can try out, visualize, and practice what they just learned. Explanations will be more detailed and in-depth, and the learning experience will be guided, in contrast to the free usage of the sandbox environment.

Overall our product seeks to cater to those who are interested in learning about neural networks, from those who seek an in-depth learning experience to those who are just curious and wish to freely experiment with neurons and layers.