Predictive Analytics for Complex IoT Data Streams

Abstract

Abstract-The requirements of analyzing heterogeneous data streams and detecting complex patterns in near real-time have raised the prospect of Complex Event Processing (CEP) for many internet of things (IoT) applications. Although CEP provides a scalable and distributed solution for analyzing complex data streams on the fly, it is designed for reactive applications as CEP acts on near real-time data and does not exploit historical data. In this regard, we propose a proactive architecture which exploits historical data using machine learning (ML) for prediction in conjunction with CEP. We propose an adaptive prediction algorithm called Adaptive Moving Window Regression (AMWR) for dynamic IoT data and evaluated it using a real-world use case with an accuracy of over 96%. It can perform accurate predictions in near real-time due to reduced complexity and can work along CEP in our architecture. We implemented our proposed architecture using open source components which are optimized for big data applications and validated it on a use-case from Intelligent Transportation Systems (ITS). Our proposed architecture is reliable and can be used across different fields in order to predict complex events.

An Ingestion and Analytics Architecture for IoT Applied to Smart City Use Cases

“…Our modular approach enables exploration of other unsupervised or supervised methods for the same problem. In addition, our architecture can be used for additional applications; for example, one can train regression models with Spark MLlib using Madrid Council's historical dataset and provide traffic predictions [33].…”

Soccer Event Detection

“…Data are playing an increasingly key role in sports, but they must be processed to extract meaningful information [2,3]. Data-driven decision plays a significant role in soccer and many other sports.…”

Visual Reasoning on Complex Events in Soccer Videos Using Answer Set Programming

“…Axiom (1) and (2) state that a fluent is initiated with the occurrence of an event that initiates it, and that fluent will be terminated when another event occurs and terminates it. Axiom (3) states that if a fluent holds at time-point T and is not terminated in T , then the fluent is true at the next time-point T 1 . Axiom (5) states that if a fluent is initiated by some event that occurs at time-point T , then the fluent is true at T 1 .…”

“…The increase in availability of data in both structured and unstructured formats is a common trend nowadays: on the other hand, information extraction for a meaningful use from this ocean of data is still a challenging task. The interpretation of these data need to be automated in order to be transformed into operational knowledge [3,20]. In particular, events are mostly important pieces of such knowledge, as they represent activities and happenings in the represented scenario.…”