https://github.com/kentonishi/simple-neural-network
Simple neural network implementation in C++.
https://github.com/kentonishi/simple-neural-network
Last synced: 4 months ago
JSON representation
Simple neural network implementation in C++.
- Host: GitHub
- URL: https://github.com/kentonishi/simple-neural-network
- Owner: KentoNishi
- Created: 2019-03-11T05:05:01.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2019-04-30T17:09:50.000Z (about 6 years ago)
- Last Synced: 2025-01-03T10:11:41.871Z (5 months ago)
- Language: C++
- Homepage:
- Size: 27.9 MB
- Stars: 1
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Simple-Neural-Network-v3
Inclusion of biases, on top of the [Simple Neural Network v2](https://github.com/KentoNishi/Simple-Neural-Network/tree/v2) project.
## Download
Clone the repository with git.
```
git clone https://github.com/KentoNishi/Simple-Neural-Network.git
```
Switch to the v3 branch.
```
git checkout v3
```
## Files
``main.cpp`` in the root directory contains the basic code for the neural network.
## Breakdown
### Basic Network
The ``main`` function serves the following purposes.
```cpp
int main(){ // function definition
Network network=Network(); // create network
vector samples; // list of samples
set classSet; // a set (ordered list with no duplicates) of output types
// custom datasets can be created in the following loop.
for(int i=0;i<10;i++){ // generate 10 samples
int a=rand()%2; // random number a
int b=rand()%2; // random number b
int c=(a&b); // binary and operation
Sample sample=Sample({float(a),float(b)},c); // list of float input values, one output value
samples.push_back(sample); // push back sample to list
}
... // data formatting for processing
// layer configurations can be specified by a vector of integers.
// layer configurations do not include the input and output layers.
network.init(samples,{5,5}); // initialize the network with the vector of samples and a layer configuration
network.train(0.1,0.001); // Specify a learning rate and minimum error value.
return 0; // exit the program
} // end function
```
## Execution
Compile and run the desired program.
```
g++ main.cpp -o result;./result;rm result;
```
## Other Information
* To learn more about neural networks, check out [this article](https://mattmazur.com/2015/03/17/a-step-by-step-backpropagation-example/) about backpropagation.