Welcome!

Microservices Expo Authors: Elizabeth White, Gregor Petri, Gerardo A Dada, Roger Strukhoff, XebiaLabs Blog

Related Topics: Microservices Expo, Java IoT, Industrial IoT, Microsoft Cloud, IoT User Interface, Apache

Microservices Expo: Article

Intelligent Complex Event Processing with Artificial Neural Network

Solve highly complex problems in real or near real time

In the current world, data is continuously being generated across various layers of organizations and environment due to changes in the system states or due to the occurrence of new events. These changes in the state of the existing system can happen due to the arrival of a new order request, customer service calls for complaints or feedback, changes in the company stock prices, text or multimedia messages, emails, social media posts, traffic reports, weather reports or any other kind of data. Simply producing reports using these data on a pre-defined schedule is not enough. Decision makers need real-time alerts and intelligent insight of all that is happening within and around the organization so that they may take meaningful reactive and proactive action before it is too late based on the new information being continuously generated.

A powerful technique called Complex Event Processing (CEP) is used for analyzing events coming from multiple sources over a specific period of time by detecting complex patterns between events and by making correlations. Apart from CEP, Artificial Neural Network (ANN) is also used to model complex relationships between input events data. Both the approaches have their own pros and cons. In this article, we tried to describe a use case in the health care domain with the solution architecture using both CEP and ANN, combining the best capabilities of both the approaches. We have shown how one can use both the techniques together to solve highly complex problems in real or near real time.

The following two sections gives brief introduction about CEP and ANN respectively with their key benefits. In section 4, we have explained the approach which combines both the CEP and the ANN efficiently to provide better solution of complex problems. Section 5 and 6 explains the Health Care: Patient Monitoring System use case with the problem description and proposed solution approach using CEP and ANN, followed by the section with summary and conclusion.

Complex Event Processing
Complex event processing is one of the key Operational Intelligence technology used to process one or more stream of data and information (also known as events) and deriving a meaningful conclusion using them. It allows one to set the request for an analysis or some query and then have it continuously executed and evaluated over time against one or many streams of events in a highly efficient manner. CEP is all about processing events that combines data from many sources to infer events or patterns that suggest more complicated circumstances [1].  For example, CEP can be used as Fraud Detection system, to detect suspicious credit card usage by monitoring credit card activity in real time and relating the current transactions with the historical data about a particular customer. The historical data which can be used by CEP Fraud Detection system can be an average transaction amount, minimum and maximum values of the previous transactions, transaction frequencies, locality etc. On detecting fraudulent activity, CEP system can send an alert via an SMS or email to the customer or the credit card service provider to take quick reaction.

The primary goal of CEP is to (1) detect meaningful events or pattern of events which signifies either threats or opportunities from the series of events being received continuously and (2) send alerts for the same to responsible entity to respond as quickly as possible. The following diagram (as figure-1) describes high level view of the CEP system.

Figure 1: High-level view of the CEP system

As shown in Figure 1, the core of the complex event processing system is made up of set of input adapters, set of output adapters and various event processing modules such as event filtering modules, in-memory caching, aggregation over different windows (time-window, sliding window, tumbling window etc.), database lookups module, database writes module, correlation, joins, event pattern matching, state machines, dynamic queries etc. More the number of I/O adapters supported by the CEP, more flexible and adaptable it is and will be able to cover wide range of use cases as compared to the CEP tool having support for limited set of I/O adapters.

Key Benefits of CEP
The following are some of the key benefits the CEP provides to the business.

  • Automatically identifies rare but important relationships between seemingly unrelated events or stream of events and accelerate timely responses to both the threats and opportunities.
  • Using sophisticated analysis and event pattern matching techniques, the CEP improves resource allocation and timely problem resolution by prioritize situations that require the most urgent attention in real or near real time based on arrival of events.
  • CEP helps organization to reduce operating costs by monitoring end-to-end performance of the system and provide timely alerts to rapidly identify potential SLA violations.
  • CEP helps organization to fine tune their business processes by correlating SLA performance with industry metrics e.g. Six Sigma and various Quality metrics, to enhance overall productivity.

Artificial Neural Network
An Artificial Neural Network (ANN) is a computational model which resembles with the way human brain is made up of in structure and the way it works. Similar to human brain which is made up of billions of neurons interconnected by synapses, the ANN can be form as a network of computational nodes connected with each other through links. The ANN needs to be trained repeatedly with specific set of training data before it can be used in production environment. Due to its adaptive nature, the internal structure of the ANN can easily be changed based on external or internal information that flows through the network during the learning phase [2]. The links are assigned weights during training process, which regulate the flow of data from one node to another. ANNs are used to model complex relationships between inputs and outputs data. ANN can efficiently find various patterns in input data or to predict future values of the system parameters. Due to its flexible construct, ANN can be very helpful in modeling complex systems which are very difficult otherwise by using traditional modeling techniques. Artificial neural networks are being applied in diverse of domains and fields. They are extensively used for doing image processing and recognition, speech recognition, credit card fraud detection, for prediction of protein structure in biotechnology and in the field of genetic science.

Artificial neural network consists of two types of interfaces with the external world, the input and the output. Since the ANN is made up of nodes or neurons and the links between them, a subset of total nodes in the ANN act as input nodes, which take data from the external world, a subset of nodes act as output node, which produces result and zero or more hidden nodes act as intermediary nodes, with having only connections with input or output nodes or other hidden nodes.  Hence, the ANN is made up of nodes in input layer, nodes in output layer and zero or more internal layers.

Figure 2: High-level view of artificial neural network

The high level view of ANN is shown in figure-2. The diagram shows a typical neural network with total 12 nodes, three nodes in the input layer, seven nodes in the hidden layer and two nodes in the output layer. Before the neural network can be used in actual production environment, it is needed to be trained for particular environment. The process of training of ANN is called learning of neural network, which is generally done in one of the following three ways:  (a) supervised learning; (b) unsupervised learning and (c) reinforcement learning. The more details about the ANN learning can be found in [2].

Key Benefits of ANN
Since ANNs can infer a function from inputs, they particularly are used in the applications where the complexity of the input data or system modeling makes the design of such a function impractical using traditional approaches. Following are some of the key benefits ANN provides.

  • It is very easy to apply ANN to problem domains where the relationships are quite dynamic or non-linear among the input and output.
  • Since ANN is capable of capturing many kind of relationships and complex patterns among data, ANN allows user to easily model the system which otherwise is very difficult or impossible to represent through traditional modeling approaches.
  • The training information is not stored in any single element but is distributed in the entire network structure. This makes ANN fault tolerant and it reduces the impact of erroneous input on the result.

CEP and ANN Together
Having seen the key properties and benefits of using both, CEP and ANN, this section describes what if one apply both together for specific set of problems to make the modeling of the system and solution easy and efficient. The CEP is best in accepting data or events from multiple channels and apply various event processing operations on it, such as event filtering, event pattern matching, aggregation etc. Apart from that user can configure alerts based on various thresholds on various system parameters. But the CEP tools lakes the ability to predict future events or determine the values of the system parameters for future events, which can be efficiently done by the ANN. So if we combine best of CEP and best of ANN for a particular problem, the resulting solution could be very effective and efficient. In the following sections, we have described how the CEP and the ANN can be used together to solve a particular problem of patient monitoring system in the domain of Health care and medicines.

Patient Monitoring System
The patient monitoring system monitors and keeps track of various body parameters of the patient and provides the data for analysis to monitoring system. Various body parameters could be blood pressure, the percentage of oxygen in the blood, glucose level in the blood, heart beat rate, change in body temperature etc. Data provided by the patient monitoring system helps to make diagnostic decisions easy and more reliable. The quality of patient treatment and care giving can greatly be improved with the use of patient monitoring systems, since it allows generating alerts in case of sudden changes in the patient body parameters which could be dangerous to the patient's health or could be life threatening some time [3].

A Use Case
Goals of the patient monitoring system are to (1) continuously keeps track of the patient's body parameters and store the data for present or future references, (2) identify life-threatening changes in patient's body and raises timely alarms for the same, and (3) to determine whether patient's health is in normal condition or it is improving or worsening based on the continuously arriving input data from various medical monitors. Since no two human bodies react in a same way against given situation or medication, it is very difficult to derived common rule set which can be applied to all human bodies. Similarly, one person's body also reacts differently in different medical and environmental situations. For example, a particular heart beat rate can be normal in some situation, while the same can be very abnormal in the other situation. So to judge the proper health condition, a trained professional is required, i.e. a specialist doctor, who studies all the observations and determine the correct state of patient's health. If the patient monitoring system is equipped with some intelligent agent who will use patient's medical history and current body parameters observations, then quality of patient care delivery can greatly be improved. We combine CEP and ANN together to propose system architecture which tries to act as an intelligent agent of the patient monitoring system, which is described in the following section.

System Architecture of the intelligent patient monitoring system using CEP and ANN
The following diagram, in Figure 3, shows the architecture of the intelligent patient monitoring system using CEP and ANN. There are total five key components; (1) Medical monitors, (2) CEP, (3) Patient's medical history and diagnosis data store, (4) ANN and (5) ANN output to action message converter.

(1) Medical Monitors
Medical monitors are medical devices used for monitoring patient's body parameters. It can consist of one or more body parameter sensors, processing components, display devices as well as communication links for displaying, recording or transmitting data or results elsewhere through a monitoring network. In the proposed architecture, the data generated by medical monitors are fed into the CEP system. [3]

Figure 3: Architecture of the intelligent patient monitoring system using CEP and ANN

The CEP section of the proposed architecture is one of the key components of the system. It receives all the monitored data and applies various event processing techniques, such as filtering, aggregation etc. over input event streams and provides the data for further processing to ANN module. Various input adapters available in CEP make it possible to collect data from different types of sensors or monitors and process them collectively. In CEP module, various event processing rule are written specific to the patient.

(3) Patient's medical history and diagnosis data store
This is the data store where patient's medical history and diagnosis data is stored. It could be traditional RDBMS storage system. The data stored in this storage are used for ANN training purpose. The new data is continuously added into the same data storage and will be used next time when ANN will be trained again with patient's latest medical and diagnosis data.

(4) ANN
The ANN model for the patient is computational neural network specific to the patient and trained using patient's all medical and diagnosis data. This trained ANN model is used for real-time diagnosis and care delivery. The decision is taken based on the input data coming from the CEP output adapters. The patient specific ANN model is trained at regular interval may be daily or on need bases. These regular updates which include latest knowledge about measured body parameters, diagnosis and medication information of the patient, helps ANN model to make accurate predictions. It is also possible to make ANN take biased decision by giving more weight to either historical data or the latest data during training. All these make ANN the most critical component of the system.

(5) ANN output to action message converter
The output generated by the ANN is generally real numbers and they are needed to be mapped to the meaningful information so that appropriate action can be taken. This is done by the ANN output to action message converter. The module not only map ANN output to real world information but it can also sends action data or alerts to devices or human being through email, SMS, alarm system etc. The threshold for various alerts can be configured so it can adapt to the changes happening to the health and body.

Together all these components make a very flexible, intelligent and efficient patient monitoring system. The proposed architecture shows how one can use CEP and ANN together more effectively to model the complex problem and provide efficient solution alternative over the traditional approaches.

Conclusion
Complex event processing and artificial neural network are the two widely used solution techniques for the problems that are very difficult to model using traditional approaches. In this article, we have described both the approaches in brief with their key capabilities. We have also described a use case for intelligent patient monitoring system with the solution architecture using both CEP and ANN and combining the best capabilities of both the approaches. We have shown how one can use both the techniques together to solve highly complex problems in real or near real time.

References

  1. Complex event processing, http://en.wikipedia.org/wiki/Complex_event_processing#cite_note-1
  2. Artificial neural network, http://en.wikipedia.org/wiki/Artificial_neural_network
  3. Patient Monitoring Systems - Part 1, http://www.philblock.info/hitkb/p/patient_monitoring_systems.html

More Stories By Kamalkumar Mistry

Kamalkumar Mistry is a Technology Analyst at Infosys Limited, Pune, India. At Infosys, he is part of a research group called Infosys Labs (http://www.infosys.com/infosys-labs).

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.


@MicroservicesExpo Stories
Kubernetes is a new and revolutionary open-sourced system for managing containers across multiple hosts in a cluster. Ansible is a simple IT automation tool for just about any requirement for reproducible environments. In his session at @DevOpsSummit at 18th Cloud Expo, Patrick Galbraith, a principal engineer at HPE, discussed how to build a fully functional Kubernetes cluster on a number of virtual machines or bare-metal hosts. Also included will be a brief demonstration of running a Galera MyS...
About a year ago we tuned into “the need for speed” and how a concept like "serverless computing” was increasingly catering to this. We are now a year further and the term “serverless” is taking on unexpected proportions. With some even seeing it as the successor to cloud in general or at least as a successor to the clouds’ poorer cousin in terms of revenue, hype and adoption: PaaS. The question we need to ask is whether this constitutes an example of Hype Hopping: to effortlessly pivot to the ...
Today’s IT environments are increasingly heterogeneous, with Linux, Java, Oracle and MySQL considered nearly as common as traditional Windows environments. In many cases, these platforms have been integrated into an organization’s Windows-based IT department by way of an acquisition of a company that leverages one of those platforms. In other cases, the applications may have been part of the IT department for years, but managed by a separate department or singular administrator. Still, whether...
As we enter the final week before the 19th International Cloud Expo | @ThingsExpo in Santa Clara, CA, it's time for me to reflect on six big topics that will be important during the show. Hybrid Cloud: This general-purpose term seems to provide a comfort zone for many enterprise IT managers. It sounds reassuring to be able to work with one of the major public-cloud providers like AWS or Microsoft Azure while still maintaining an on-site presence.
I’m a huge fan of open source DevOps tools. I’m also a huge fan of scaling open source tools for the enterprise. But having talked with my fair share of companies over the years, one important thing I’ve learned is that you can’t scale your release process using open source tools alone. They simply require too much scripting and maintenance when used that way. Scripting may be fine for smaller organizations, but it’s not ok in an enterprise environment that includes many independent teams and to...
Get deep visibility into the performance of your databases and expert advice for performance optimization and tuning. You can't get application performance without database performance. Give everyone on the team a comprehensive view of how every aspect of the system affects performance across SQL database operations, host server and OS, virtualization resources and storage I/O. Quickly find bottlenecks and troubleshoot complex problems.
Between 2005 and 2020, data volumes will grow by a factor of 300 – enough data to stack CDs from the earth to the moon 162 times. This has come to be known as the ‘big data’ phenomenon. Unfortunately, traditional approaches to handling, storing and analyzing data aren’t adequate at this scale: they’re too costly, slow and physically cumbersome to keep up. Fortunately, in response a new breed of technology has emerged that is cheaper, faster and more scalable. Yet, in meeting these new needs they...
In his general session at 19th Cloud Expo, Manish Dixit, VP of Product and Engineering at Dice, discussed how Dice leverages data insights and tools to help both tech professionals and recruiters better understand how skills relate to each other and which skills are in high demand using interactive visualizations and salary indicator tools to maximize earning potential. Manish Dixit is VP of Product and Engineering at Dice. As the leader of the Product, Engineering and Data Sciences team at D...
In IT, we sometimes coin terms for things before we know exactly what they are and how they’ll be used. The resulting terms may capture a common set of aspirations and goals – as “cloud” did broadly for on-demand, self-service, and flexible computing. But such a term can also lump together diverse and even competing practices, technologies, and priorities to the point where important distinctions are glossed over and lost.
Monitoring of Docker environments is challenging. Why? Because each container typically runs a single process, has its own environment, utilizes virtual networks, or has various methods of managing storage. Traditional monitoring solutions take metrics from each server and applications they run. These servers and applications running on them are typically very static, with very long uptimes. Docker deployments are different: a set of containers may run many applications, all sharing the resource...
The 20th International Cloud Expo has announced that its Call for Papers is open. Cloud Expo, to be held June 6-8, 2017, at the Javits Center in New York City, brings together Cloud Computing, Big Data, Internet of Things, DevOps, Containers, Microservices and WebRTC to one location. With cloud computing driving a higher percentage of enterprise IT budgets every year, it becomes increasingly important to plant your flag in this fast-expanding business opportunity. Submit your speaking proposal ...
You have great SaaS business app ideas. You want to turn your idea quickly into a functional and engaging proof of concept. You need to be able to modify it to meet customers' needs, and you need to deliver a complete and secure SaaS application. How could you achieve all the above and yet avoid unforeseen IT requirements that add unnecessary cost and complexity? You also want your app to be responsive in any device at any time. In his session at 19th Cloud Expo, Mark Allen, General Manager of...
In his session at 19th Cloud Expo, Claude Remillard, Principal Program Manager in Developer Division at Microsoft, contrasted how his team used config as code and immutable patterns for continuous delivery of microservices and apps to the cloud. He showed how the immutable patterns helps developers do away with most of the complexity of config as code-enabling scenarios such as rollback, zero downtime upgrades with far greater simplicity. He also demoed building immutable pipelines in the cloud ...
"Dice has been around for the last 20 years. We have been helping tech professionals find new jobs and career opportunities," explained Manish Dixit, VP of Product and Engineering at Dice, in this SYS-CON.tv interview at 19th Cloud Expo, held November 1-3, 2016, at the Santa Clara Convention Center in Santa Clara, CA.
@DevOpsSummit taking place June 6-8, 2017 at Javits Center, New York City, is co-located with the 20th International Cloud Expo and will feature technical sessions from a rock star conference faculty and the leading industry players in the world. @DevOpsSummit at Cloud Expo New York Call for Papers is now open.
Rapid innovation, changing business landscapes, and new IT demands force businesses to make changes quickly. In the eyes of many, containers are at the brink of becoming a pervasive technology in enterprise IT to accelerate application delivery. In this presentation, attendees learned about the: The transformation of IT to a DevOps, microservices, and container-based architecture What are containers and how DevOps practices can operate in a container-based environment A demonstration of how ...
Cloud Expo, Inc. has announced today that Andi Mann returns to 'DevOps at Cloud Expo 2017' as Conference Chair The @DevOpsSummit at Cloud Expo will take place on June 6-8, 2017, at the Javits Center in New York City, NY. "DevOps is set to be one of the most profound disruptions to hit IT in decades," said Andi Mann. "It is a natural extension of cloud computing, and I have seen both firsthand and in independent research the fantastic results DevOps delivers. So I am excited to help the great t...
Without lifecycle traceability and visibility across the tool chain, stakeholders from Planning-to-Ops have limited insight and answers to who, what, when, why and how across the DevOps lifecycle. This impacts the ability to deliver high quality software at the needed velocity to drive positive business outcomes. In his general session at @DevOpsSummit at 19th Cloud Expo, Phil Hombledal, Solution Architect at CollabNet, discussed how customers are able to achieve a level of transparency that e...
Application transformation and DevOps practices are two sides of the same coin. Enterprises that want to capture value faster, need to deliver value faster – time value of money principle. To do that enterprises need to build cloud-native apps as microservices by empowering teams to build, ship, and run in production. In his session at @DevOpsSummit at 19th Cloud Expo, Neil Gehani, senior product manager at HPE, discussed what every business should plan for how to structure their teams to delive...
DevOps is being widely accepted (if not fully adopted) as essential in enterprise IT. But as Enterprise DevOps gains maturity, expands scope, and increases velocity, the need for data-driven decisions across teams becomes more acute. DevOps teams in any modern business must wrangle the ‘digital exhaust’ from the delivery toolchain, "pervasive" and "cognitive" computing, APIs and services, mobile devices and applications, the Internet of Things, and now even blockchain. In this power panel at @...