mirror of
https://github.com/codeplea/genann
synced 2024-11-21 22:11:34 +03:00
Added to documentation.
This commit is contained in:
parent
064703e332
commit
d9da5edab4
35
README.md
35
README.md
@ -2,9 +2,10 @@
|
|||||||
|
|
||||||
#Genann
|
#Genann
|
||||||
|
|
||||||
Genann is a very minimal library for training and using feedforward artificial neural
|
Genann is a minimal, but well tested and documented, library for training and
|
||||||
networks (ANN) in C. Its primary focus is on being simple, fast, and hackable. It achieves
|
using feedforward artificial neural networks (ANN) in C. Its primary focus is
|
||||||
this by providing only the necessary functions and little extra.
|
on being simple, fast, reliable, and hackable. It achieves this by providing only the
|
||||||
|
necessary functions and little extra.
|
||||||
|
|
||||||
##Features
|
##Features
|
||||||
|
|
||||||
@ -14,13 +15,13 @@ this by providing only the necessary functions and little extra.
|
|||||||
- Fast and thread-safe.
|
- Fast and thread-safe.
|
||||||
- Easily extendible.
|
- Easily extendible.
|
||||||
- Implements backpropagation training.
|
- Implements backpropagation training.
|
||||||
- Compatible with training by alternative methods (classic optimization, genetic algorithms, etc)
|
- *Compatible with alternative training methods* (classic optimization, genetic algorithms, etc)
|
||||||
- Includes examples and test suite.
|
- Includes examples and test suite.
|
||||||
- Released under the zlib license - free for nearly any use.
|
- Released under the zlib license - free for nearly any use.
|
||||||
|
|
||||||
##Example Code
|
##Example Code
|
||||||
|
|
||||||
Four example programs are included.
|
Four example programs are included with the source code.
|
||||||
|
|
||||||
- `example1.c` - Trains an ANN on the XOR function using backpropagation.
|
- `example1.c` - Trains an ANN on the XOR function using backpropagation.
|
||||||
- `example2.c` - Trains an ANN on the XOR function using random search.
|
- `example2.c` - Trains an ANN on the XOR function using random search.
|
||||||
@ -29,8 +30,13 @@ Four example programs are included.
|
|||||||
|
|
||||||
##Quick Example
|
##Quick Example
|
||||||
|
|
||||||
Here we create an ANN, train it on a set of labeled data using backpropagation,
|
We create an ANN taking 2 inputs, having 1 layer of 3 hidden neurons, and
|
||||||
ask it to predict on a test data point, and then free it:
|
providing 2 outputs. It has the following structure:
|
||||||
|
|
||||||
|
![NN Example Structure](./doc/e1.png)
|
||||||
|
|
||||||
|
We then train it on a set of labeled data using backpropagation and ask it to
|
||||||
|
predict on a test data point:
|
||||||
|
|
||||||
```C
|
```C
|
||||||
#include "genann.h"
|
#include "genann.h"
|
||||||
@ -38,10 +44,10 @@ ask it to predict on a test data point, and then free it:
|
|||||||
/* Not shown, loading your training and test data. */
|
/* Not shown, loading your training and test data. */
|
||||||
double **training_data_input, **training_data_output, **test_data_input;
|
double **training_data_input, **training_data_output, **test_data_input;
|
||||||
|
|
||||||
/* New network with 5 inputs,
|
/* New network with 2 inputs,
|
||||||
* 2 hidden layer of 10 neurons each,
|
* 1 hidden layer of 3 neurons each,
|
||||||
* and 1 output. */
|
* and 2 outputs. */
|
||||||
genann *ann = genann_init(5, 2, 10, 1);
|
genann *ann = genann_init(2, 1, 3, 2);
|
||||||
|
|
||||||
/* Learn on the training set. */
|
/* Learn on the training set. */
|
||||||
for (i = 0; i < 300; ++i) {
|
for (i = 0; i < 300; ++i) {
|
||||||
@ -50,7 +56,8 @@ for (i = 0; i < 300; ++i) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
/* Run the network and see what it predicts. */
|
/* Run the network and see what it predicts. */
|
||||||
printf("Output for the first test data point is: %f\n", *genann_run(ann, test_data_input[0]));
|
double const *prediction = genann_run(ann, test_data_input[0]);
|
||||||
|
printf("Output for the first test data point is: %f, %f\n", prediction[0], prediction[1]);
|
||||||
|
|
||||||
genann_free(ann);
|
genann_free(ann);
|
||||||
```
|
```
|
||||||
@ -86,7 +93,7 @@ void genann_train(genann const *ann, double const *inputs,
|
|||||||
```
|
```
|
||||||
|
|
||||||
`genann_train()` will preform one update using standard backpropogation. It
|
`genann_train()` will preform one update using standard backpropogation. It
|
||||||
should be called by passing in an array of inputs, an array of expected output,
|
should be called by passing in an array of inputs, an array of expected outputs,
|
||||||
and a learning rate. See *example1.c* for an example of learning with
|
and a learning rate. See *example1.c* for an example of learning with
|
||||||
backpropogation.
|
backpropogation.
|
||||||
|
|
||||||
@ -133,7 +140,7 @@ FAQ](http://www.faqs.org/faqs/ai-faq/neural-nets/part1/) is an excellent
|
|||||||
resource for an introduction to artificial neural networks.
|
resource for an introduction to artificial neural networks.
|
||||||
|
|
||||||
If you're looking for a heavier, more opinionated neural network library in C,
|
If you're looking for a heavier, more opinionated neural network library in C,
|
||||||
I highly recommend the [FANN library](http://leenissen.dk/fann/wp/). Another
|
I recommend the [FANN library](http://leenissen.dk/fann/wp/). Another
|
||||||
good library is Peter van Rossum's [Lightweight Neural
|
good library is Peter van Rossum's [Lightweight Neural
|
||||||
Network](http://lwneuralnet.sourceforge.net/), which despite its name, is
|
Network](http://lwneuralnet.sourceforge.net/), which despite its name, is
|
||||||
heavier and has more features than Genann.
|
heavier and has more features than Genann.
|
||||||
|
9
doc/e1.dot
Normal file
9
doc/e1.dot
Normal file
@ -0,0 +1,9 @@
|
|||||||
|
digraph G {
|
||||||
|
rankdir=LR;
|
||||||
|
|
||||||
|
{i1 i2} -> {h1 h2 h3} -> {o1 o2};
|
||||||
|
i1, i2, h1, h2, h3, o1, o2 [shape=circle; label="";];
|
||||||
|
|
||||||
|
input -> hidden -> output [style=invis;];
|
||||||
|
input, hidden, output [shape=plaintext;];
|
||||||
|
}
|
BIN
doc/e1.png
Normal file
BIN
doc/e1.png
Normal file
Binary file not shown.
After Width: | Height: | Size: 22 KiB |
Loading…
Reference in New Issue
Block a user