H2O.ai has announced Wave ML’s introduction, an open-source automatic machine learning package to build and integrate predictive AI/ML models into Wave apps. By providing a simple, high-level API for training, deploying, scoring and explaining machine learning models, Wave ML, let’s users rapidly build and deploy interactive predictive and decision-support applications over the web.
With the number of data scientists worldwide increasing at a tremendous speed, the number of developers is also surging who are building applications for various business needs. According to H2O.ai, currently, there are approximately 23 million Python developers globally, of which many are not proficient with data science. And that’s where H2O’s Wave ML abstract away the complexity of machine learning and empower developers to solve business needs in their applications with the power of AutoML.
According to the official blog post, Wave ML uses the open-source H2O AutoML under the hood and is designed to transparently switch over to Driverless AI when deployed on H2O.ai Hybrid Cloud.
The blog post further stated that Wave, Wave ML and H2O AutoML are all 100% open-source under Apache v2 to build and deploy predictive apps in their preferred deployment environment.
Considering H2O Wave ML is a companion Python package to H2O Wave, both are available on PyPI and can be installed in tandem using pip:
Get to know more here.
Besides, Wave ML provides four high-level functions — train a model on a dataset, given the column to be predicted; make a prediction; save the model; load the previously saved model.
Thus, with Wave ML, developers can now harness the power of the H2O platform and integrate predictive models into their Wave apps, with just a couple of code lines.
H2O.ai has also announced that it is further working on automating the package, which will let the developers code-generate complete, interactive apps.
Read the entire blog here.
