Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
This is a configurable Neural Network written in C#. The Network functionality is completely decoupled from the UI and can be ported to any project. You can also export and import fully trained networks and datasets.
| Date | Stars |
|---|---|
| 2026-07-31 | 281 |
| 2026-08-01 | 281 |
| 2026-08-02 | 281 |
| 2026-08-06 | 281 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Neural Networks Introduction ------------ If this is your first foray into Neural Networks, welcome! I hope you enjoy yourself as much as I have. This project is an attempt at creating an application that allows for quick interactions with a basic neural network. This project is written in C# and uses C# 6.0 Syntax. You will need an environment that is capable of compiling the C# 6.0 syntax in order to use this program.  What is a Neural Network? ----- Great question! A Neural Network can be thought of as a series of nodes (or neurons) that are interconnected, much like they are in the brain. The network can have any number (N) of inputs and any number (M) of outputs. In between the inputs and outputs are a series of "hidden" neurons that make up the hidden layers of the network. These hidden layers provide the meat of the network and allow for some of the neat functionalities we can get out of a Neural Network. What are the Parts of a Neural Network? -------- Before explaining the pieces of a neural network, it might be helpful to start with an example. Building off of [this excellent article from 2013](http://www.codingvision.net/miscellaneous/c-backpropagation-tutorial-xor), let's use the concept of Exclusive Or (XOR). XOR will output true when the inputs differ: | Input A | Input B | Output | |:-----:|:---------:|:-----:| | false | false | false | | false | true | true | | true | false | true | | true | true | false | Considering this, let's break down a Neural Network into its three basic parts: 1. The Inputs * These are the inputs into the Neural Network. From the XOR example above, the inputs would be Input A and Input B. * Each input can be considered a Neuron whose output is the initial input value. 2. The Hidden Layers * This is the meat of the Neural Network. This is where the magic happens. The Neurons in this layer are assigned weights for each of their inputs. These weights start off fairly random, but as the network is "trained" (discussed below), the weights are adjusted in order to make the neuron's output, and therefore the Neural Network's output closer to the expected result. 3. The Outputs * These are the outputs from the system. From the XOR example above, the output from the system would be either 'true' or 'false'. In the Neural Network, the Outputs are the last line of Neurons. These Neurons are also assigned a weight for each of their inputs and are "fed" by the Neurons in the hidden layer. Using the XOR example, if we were to give our Neural Network the inputs 'true' and 'false' we would expect the system to return 'true'. How Does it Work? ------- Because I love examples, here's another: | Input A | Input B | Input C | Output | |:-----:|:---------:|:-----:|:------| | true | false | false | true | | false | true | true | false | | true | fase | true | false | | true | true | true | true | In the above table, we can infer the following patterns: 1. The output is true if the number of inputs set to true is odd OR the number of inputs set to false is even. 2. The output is false if the number of inputs set to true is even OR the number of inputs set to false is odd. The job of the Neural Network is to try and figure out that pattern. It does this via training. #### How Do We Train the Neural Network? Training the Neural Network is accomplished by giving it a set of input data and the expected results for those inputs. This data is then continuously run through the Neural Network until we can be reasonably sure that it has a grasp of the patterns present in that data. In this project, the Neural Network is trained via the following common Neural Network training methods: 1. Back-Propagation * After each set of inputs is run through the system and an output generated, that output is validated against the expected output. * The percentage of error that results is then propagated backwards (hence
Excerpt of 13,480 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:8d0c132b1ee53851, llm:Description: "This is a configurable Neural Network written in C#. The Network functionality is completely decoupled from the UI ... export and import fully trained networks and datasets."
matched fp:8d0c132b1ee53851, llm:Description: "This is a configurable Neural Network written in C#. The Network functionality is completely decoupled from the UI ... export and import fully trained networks and datasets."
matched fp:8d0c132b1ee53851, llm:Description: "This is a configurable Neural Network written in C#. The Network functionality is completely decoupled from the UI ... export and import fully trained networks and datasets."