Conclusion

Data Science and Machine Learning

The tutorial is now ended. We have built together a building energy consumption prediction system that exposes its predictions in REST data service.

We have begun the journey from importing raw data sources, then explored them in the business-oriented Data Catalog and the analytics notebook-based environment Apache Zeppelin for Denodo.

Later, we have logically combined the data in the Web Design Studio, and after training and saving the model again in Apache Zeppelin for Denodo, we have published the REST data service that enables the consumption of the predictions, in real-time from any application.

We hope that through this hands-on real-world example we were able to clearly show how the Denodo Platform brings agility and value to all stages of machine learning-focused projects and we have enabled you with the necessary background and knowledge to start thinking about data virtualization implementations in other analytical initiatives.