AutoML, AI tool

[EdEarl]

Quote

http://www.ml4aad.org/automl/

Machine learning (ML) has achieved considerable successes in recent years and an ever-growing number of disciplines rely on it. However, this success crucially relies on human machine learning experts to perform the following tasks:

• Preprocess the data
• Select appropriate features
• Select an appropriate model family
• Optimize model hyperparameters
• Postprocess machine learning models
• Critically analyze the results obtained.

As the complexity of these tasks is often beyond non-ML-experts, the rapid growth of machine learning applications has created a demand for off-the-shelf machine learning methods that can be used easily and without expert knowledge. We call the resulting research area that targets progressive automation of machine learning AutoML.
Although it focuses on end users without expert ML knowledge, AutoML also offers new tools to machine learning experts, for example to:

• Perform architecture search over deep representations
• Analyse the importance of hyperparameters.

Following the paradigm of Programming by Optimization, AutoML advocates the development of flexible software packages that can be instantiated automatically in a data-driven way.

AutoML is the first AI automation tool that I've heard of; although, there may be others.

Previously, before the current AI epoch, Computer Aided Engineering and Design systems of various types were sold. However, few were a commercial success. The AutoML team seems to have a good technique, since they are integrating AI into the system they are designing and putting AI into the systems it produces. Any time they can save developing will be used for other things. In the limit, an AI will be able to do anything a human can. At this time this tool falls short of that goal, but I believe they will continue to improve AutoML and perhaps build other related tools.

Their design with data driven software packages means an AI could be composed of a million packages, in which case programming personnel cost would be too much unless tools can significantly reduce the workload. The important part of the human job is to lend insight and direction to the less than genus AI tools.

AI developers have an eye on the Singularity as if it were the last Nobel Prize. I think they are on the right path, and that they will succeed within 10-15 years.