This package is about Artificial Intelligence, Machine Learning and Pattern Recognition in Laravel.
The video tutorial with examples on how to use the library can be found here:
how to use AI in Laravel (full playlist)
Do AI with 2 lines of code:
$all_animals = DogsAndCats::all();
$report = RubixAi::train($all_animals, 'dog_or_cat');
that's it.
This creates a file in your laravel storage/ai_rubix/ folder that contains the model.
This model then will be used to predict:
$isDogOrCat = RubixAi::predict($needs_prediction);
echo $isDogOrCat; //prints ['dog']
composer require torian257x/ai-php-rubix-wrap-laravel
If there are any issues, please have a look at https://docs.rubixml.com/latest/installation.html in case you are trying to do something special. I recommend as well installing tensor https://github.com/Scien-ide/Tensor .
new KDNeighborsRegressor()
for regression (say estimation of price)
new KDNeighbors()
for classification (say choose dog or cat)
array_filter(
[
new NumericStringConverter(),
new MissingDataImputer(),
$needs_ohe ? new OneHotEncoder() : false,
new MinMaxNormalizer(),
]);
You can choose your own estimator if you don't like the default
Just be sure to pass it as argument to RubixAi::train(..., estimator_algorithm:<>)
Same for transformers.
You can customize the model file name as well as what attributes / columns to ignore.
This library is using a PHP standalone AI wrapper of Rubix ML