FAQ#
Here you will find answers to the most frequently asked questions.
Is Machine Learning the same thing as AI?#
Machine Learning is a subset of Artificial Intelligence (AI) that focuses on using data to train a computer to perform tasks. While machine learning (ML) has contributed substantially to the field of AI, other non-learning techniques such as rule-based or symbolic systems are also forms of artificial intelligence.
What types of problems is ML good for?#
Machine Learning is a good fit for problems in which it would be infeasible for software developers and domain experts to design a system that could encode all the necessary rulesets to obtain accurate predictions. In other words, if your problem can be solved with a few if/then statements, it is probably not a good fit for machine learning due to unnecessary complexity.
What environment (SAPI) should I run Rubix ML in?#
All Rubix ML projects are designed to run from the PHP command line interface (CLI). The reason almost always boils down to performance and memory consumption.
If you would like to serve your models in production, the preferred method is to use the Server library to spin up a high-performance standalone model server from the command line. If you plan to implement your own model server, we recommend using an asynchronous event loop such as React PHP or Swoole to prevent the model from having to be loaded into memory on each request.
To run a PHP script using the command line interface (CLI), open a terminal window and enter:
$ php example.php
Note
The PHP interpreter must be installed and in your default PATH for the above syntax to work correctly.
I'm getting out of memory errors.#
Try adjusting the memory_limit
option in your php.ini file to something more reasonable. We recommend setting this to -1
(no limit) or slightly below your device's memory supply for best results.
You can temporarily set the memory_limit
in your script by using the ini_set()
function.
ini_set('memory_limit', '-1');
Note
Training can require a lot of memory. The amount necessary will depend on the amount of training data and the size of your model. If you have more data than you can hold in memory, some learners allow you to train in batches. See the Online Learning section of the Training docs for more information.
Training is slower than usual.#
Training time depends on a number of factors including size of the dataset and complexity of the model. If you believe that training is taking unusually long then check the following factors.
- Xdebug or other debuggers are not enabled.
- You have enough RAM to hold the dataset and model in memory without swapping to disk.
Does Rubix ML support multiprocessing/multithreading?#
Yes, learners that support parallel processing (multiprocessing or multithreading) do so by utilizing a pluggable parallel computing backend such as Amp or extension such as Tensor under the hood.
Does Rubix ML support Deep Learning?#
Yes, a number of learners in the library support Deep Learning including the Multilayer Perceptron classifier and MLP Regressor.
Does Rubix ML support Reinforcement Learning?#
Not currently, but we may in the future.
Does Rubix ML support time series data?#
Yes and no. Currently, the library treats time series data like any other continuous feature. In the future, we may add algorithms that work specifically with a separate time component.
How can I contribute to the project?#
Anyone is welcome to contribute to Rubix ML. See the CONTRIBUTING guide in the project root for more info.
How can I become a sponsor?#
Check out our funding sources here and consider donating to the cause.