CLARA: Cloud Learning Autonomous Reacting Algorithms, a Machine Learning Service for the Internet of Things

The Internet of Things (IoT) is growing and it is expected that, by 2020, 31 thousand million devices will be connected. By that time, more than 7 ZetaBytes will have been sent to Cloud Services to be stored and analysed. In order to find errors, analyse and learn from all the data acquired, Machine Learning (ML) Models are being used. Cloud Service providers offer tools to simplify the development of ML Models. However, the developers still have to be involved in the development, testing, and pruning of the models. CLARA is a cloud based Machine Learning service that aims to simplify all the current developers' work by creating a data interpreter that is able to generate the best ML model for the data provided. In the following paper, research is presented that shows how CLARA can be used to receive and learn from data coming from IoT applications. Furthermore, it will show how CLARA obtains better results than other ML models trained by a developer with no theoretical background on ML.