A new Machine Learning development may change the future landscape of algorithm-building
Research in the development of Artificial Intelligence and Machine Learning has grown by leaps and bounds over the past few years, with more organisations adopting these new technologies in the running of their businesses than ever before. In fact, a recently conducted survey among top industry professionals found that AI ranks second among the most-adopted new technologies amongst businesses, one spot behind ML-based advanced analytics.
This should, of course, come as no surprise. Over the last few years, machine learning has not only grown to fulfil its potential boots but has overgrown them massively, proving to be a highly valuable computational tool in tackling a plethora of real-world problems. These include aspects like fraud detection, personalised marketing, process automation, dynamic pricing and image, audio and text classification tasks, among a multitude of others.
However, this has also given birth to a new challenge: keeping track of and quickly finding and accessing all the different machine learning algorithms being created by computer scientists all over the world to make the development of new algorithms quicker and more accessible. Keeping this in mind, researchers at the University of Cincinnati and Purdue University recently created HAMLET (Hierarchical Agent-based Machine LEarningplaTform) – “a platform that could help computer scientists and developers to browse through existing machine learning models and train or evaluate their own algorithms, thus aiding their research and development efforts.”
According to Ahmad Esmaeili, one of the researchers who carried out the study: “Organizing and keeping track of the machine learning algorithms and datasets has always been a major challenge for us, as well for as many other researchers in the field, This becomes even more critical when the number of ML solutions and components continues to grow over time and from one project to another. When developing HAMLET, we have strived to create a platform that meets the needs above by not only administering the available ML contributions and assets in a distributed way but also facilitating actions such as accessing, comparing and evaluating those resources effectively.”
The architecture is made up of a group of AI-based agents which are trained to ‘manage’ a large group of machine learning algorithms (including related resources, such as datasets and tasks that the models were trained to complete) according to a hierarchical structure based on the specific tasks they are carrying out.
HAMLET essentially starts with an empty structure that continues to grow autonomously with the introduction of new ML-based queries or resources over time. Additionally, given the fact that HAMLET can be distributed across a network of computers and devices, there is no real constraint on the size and type of algorithms and data that it can host. Its user-friendly interface facilitates a flexible query structure, and thus widens its use-case across a wide variety of tasks – such as in the training and testing of new algorithms as well.
To Be or Not to Be?
In its development, HAMLET was used to complete 120 training and four-batch testing tasks on a simulated Python-based environment. Twenty-four machine learning algorithms were tested and trained using nine renowned datasets, with results showing that it is a highly promising tool for the training and testing of ML algorithms.
Although it is still in its nascent stages, developers of HAMLET believe that it can be ameliorated in many aspects to ensure it meets current research and industrial needs. According to Esmaeili,“(the future) plan (is) to continue working on supporting more sophisticated algorithms, the survivability of the platform against failures, merging multiple platforms, and the privacy of accessing data/algorithms.”
If to be or not to be was ever a question, HAMLET is definitely to be.