The AutoML tools allow the Learning Machine to be automated so that it can be operated with minimal technical skills. Discover a selection of the top ten tools in this category.
Machine Learning offers many possibilities for companies in all industries. However, it is a complex technology that requires a deep mathematical understanding of the basis of the technology..
Fortunately, in order to democratize machine learning and to allow the greatest number of people to benefit from it, many companies are now developing and offering AutoML” type solutions to automate the Machine Learning system. Here is our selection of the best tools.
Auto-Keras, the open source AutoML tool
The Auto-Keras open source libraryThe new Machine Learning System, developed by DATA Lab at Texas A&M University, is dedicated to the automation of Machine Learning. In particular, it allows to automatically search for architectures and hyperparameters of Deep Learning models. In addition, as an open source solution, it benefits from a large community of contributors who are constantly improving it…
H20 AutoML, the AutoML for stand-alone vehicles
Thanks to the H2O AutoML interfacedevelopers can test their data for autonomous vehicles. The platform gathers models for initial testing, saving users from having to develop a model from scratch.
SMAC, the hyperparameter optimization tool
SMAC (sequential model-based algorithm configuration) is a tool dedicated to the optimization of algorithm parameters. It proves to be very efficient for the optimization of hyperparameters of Machine Learning algorithms.
AUTO-SKLEARN, the AutoML for Supervised Learning Machine
The Auto-sklearn tool is mainly dedicated to Supervised Machine Learning. It is based on the scikit-learn Machine Learning library, and allows you to automatically search for a set of data or optimize hyperparameters.
ROBO, the modular framework for Bayesian optimization
RoBo (Robust Bayesian Optimization) is a framework written in Python. It is a modular framework, allowing to easily add and exchange Bayesian optimization components such as the different acquisition functions or regression models. We find different regression models such as Bayesian neural networks or decision tree forests.
Amazon Lex, the AutoML for chatbots applications
With Amazon Lex, developers can take advantage of advanced Deep Learning features for the automatic speech recognition (ASR), speech-to-text conversion, and natural language processing. Thus, users can develop chatbots applications based on Alexa with ease.
AUTOFOLIO optimizes algorithm selection
Thanks to the configuration of algorithms, AutoFolio optimizes the performance of algorithm selection systems determining the best approach for selection and hyperparameters.
AUTOWEKA chooses the best algorithm and optimizes its hyperparameters at the same time.
Rather than perform these two tasks separately, Auto-WEKA chooses the best algorithm while optimising its hyperparameters. The tool opts for a fully automated approach, and exploits the latest innovations in the field of Bayesian optimization.
There again, Auto-PyTorch aims to optimize the choice of architecture and the configuration of the hyperparameters. Machine Learning algorithms. However, the platform stands out by opting for multi-fidelity optimization and Bayesian optimization to achieve this.
FLEXFOLIO the unified algorithm selection framework
Flexfolio is a modular architecture combining different algorithm selection techniques and approaches. It provides a unified framework for comparing and combining algorithm selection approaches.