Author Archives: admin

Posted on Author admin Categories Artificial Intelligence

Machine Learning – Evolution in making of Modern Enterprise

Those in analytics might have come across the term Machine Learning. It is often misused and glorified as a magnificent future for machines replacing humans.
Though there is some prejudice and a lot of grandiose built around ML, the fact is ML is the most powerful and advanced technology for a modern enterprise.
In a very crisp and layman language we can conclude, Machine learning can be used in automation of repetitive tasks that would otherwise need to be done manually, thereby enhance enterprise efficiency and execute repeatedly at scale.
What is Machine Learning (ML)?
ML is a specific field in computer science that emphasizes on machine programming to enhance self-performance through data and iteration.
The start point for conventional programmer and ML is almost similar – both intend to resolve problems and begin with developing familiarity with problem domain. However, the differentiation aspect is of execution. Programmers use their ingenuity to formulate a computer program to develop solution. On the other hand, data scientists who implement ML collect inputs and target values and feed the instructions to the computer to develop a program for a desired output.
Say for example, in streaming of videos, like Netflix – engineers must spend tedious hours to develop recommendation options for its users based on the history of choices or early inputs given. In certain cases, this works, the program helps pair user watched videos for recommendation based on factors such as genre. However, it is difficult for programmers to sort among a pile of thousands of titles and lakhs of subscribers with unique history of each.
While sifting through piles of data is a challenge, another problem that programmers come across is recommending videos that user might or might not like based on their watch and browse history. Chances are their interest might have changed which is quite impossible to predict.
When human intelligence fails to predict such patterns, ML pitches in. Algorithm based ML gathers data and learns from them for valid predictions rather than relying completely on human instructions. Further, ML keeps upgrading its data-based learning time to time as more and more information users provide through their browsing history.
AI v/s ML
The most common question people end-up asking is the difference or correlation between AI and ML. The answer is ML is a type of AI, a subset under the vast field of artificial intelligence. Further AI is a subset of computer science. To be precise ML entails deeper technical aspect, a specific methodology. Whereas AI is non-technical, an intelligent system that can mimic human behavior.
Supervised Learning v/s Unsupervised Learning 
Supervised learning is data mining of drawing inference from a function from labeled training data. Most of the practical machines use this format. In case of supervised learning you have input variable and an output variable, and an algorithm is used to map the function from input to output.
Subsequently, once the training process is standardized, Programmers test for program accuracy and make required amends, and repeat the entire process until they achieve full-proof accuracy in the overall process of supervised learning.
For example, Cortana and other AI enabled assistants used in your phone or other devices, is trained as an input of human voice and works as a result of this training.
In case of unsupervised learning the program is trained without the labeled and structured data. Alternatively, it means that the algorithms are trained to give results only through inputs without respective output unlike paired training in supervised learning. The algorithm learns to condition itself to process the structure of data to understand it and provide valid outputs.
Deep Learning
The way ML is a type of AI, deep learning is a subset of ML. Various streams of ML algorithms, deep learning being one of it, is related with neural network.
A neural network is based on the underlying principle of how human brain cells, the neurons, function. Its achieved by fine layers of composite units to understand and interpret correlation based on data. When the layers are deeper, hidden layers being more than one in the neural network, it’s called deep learning – it can be supervised or unsupervised, at times semi-supervised.
Engineers have already put deep learning to solve the most complicated tasks and crack the toughest ones, most critical being training self-driving cars and cancer diagnosis.
Why should enterprises explore ML?
The advancement in AI and further sophistication in ML has taken the business landscape by storm. For example, self-driven cars on the road and a weather forecasting computer program based on ML algorithms.
Machine learning as a service (MLaaS) is a set of services that provide ML tools as bundled cloud solution. MLaas is a cost-effective array that offers the enterprise advantages of ML, saving their time and exhaustive in-house establishment of ML team.
MLaaS also circumvents the infrastructure related challenges like data pre-processing, model training, model evaluation, and finally, predictions.
Some more examples include cyber security, process automation, data analysis in insurance and finance sector. It is quite possible that ML has already affected your business enterprise too in one way or the other.
The million-dollar question is how you train your teams to adapt to ML and use it successfully.
Conclusion
Automation starts with data – precisely the machine data that sits on a large assemblage of hardware, software and management tools that construct the present IT infrastructure and services. The daily addition of new devices and technology to the existing digital landscape has made the enterprise ecosystem complex. By automation of repetitive tasks and employing innovative ML, businesses can overcome the talent constraints and achieve almost zero error. In addition, automation helps gain new insights for better outcomes, drive efficiencies and improve the security features.
Machine learning will increasingly become priceless as the technology matures with time and many more businesses embrace algorithm-based learning for a smarter enterprise. The technology has already impacted most of the sectors such as car industry and insurance and finance sector which are large scale. However, there are lesser known innovations of ML too that are just about the corner, waiting to be discovered and embark on an exciting journey.
What are your thoughts on Machine Learning? We’d love to hear from you. Please post your comments.
References
http://www.ibmbigdatahub.com/blog/what-is-machine-learning
https://www.simplilearn.com/what-is-machine-learning-and-why-it-matters-article
https://blog.algorithmia.com/introduction-machine-learning-developers/

 

0
Posted on Author admin Categories Artificial Intelligence

Artificial Intelligence Transforming Financial Services

The first incidence of a machine and human crossing paths was the epic win of Deep Blue, the supercomputer, over grandmaster Garry Kasparovin 1997 in an intense game of chess. The triumph certainly triggered a debatable topic and kickstarted the rise of the machines.
In 21st century, Artificial Intelligence (AI) has moved past the celluloid and realms of research, establishing its presence in the real world amid an unparalleled convergence.
AI has been pivotal in changing the face of the entire financial services, taken by storm. Almost every financial enterprise has embraced the technology for better time keeping, cost effectiveness and value-added services.
Use of AI in financial sector has been on a surge yet at an incremental scale. In India nearly 36% of Financial institutions have already invested in AI technology and 70% are reported to do so shortly.
Development in high-density parallel processing infrastructure and an extraordinary surge in the bulk and kind of data generated, powered the adoption of Machine Learning and many more cognitive technologies. Fueling this trend is the snowball effect of cloud computing and mobility as well as open sourcing of machine learning (ML) algorithm.
Business Imperative
Financial institutions have compelling reasons to implement AI. Falling interest rates continue to impact the bottom line, banks are left with minimal options than to boost operational efficiency by minimizing the unscheduled downtime of systems and reducing the resources to meet the evolving demand.
Building on the top line through enhanced and tailored targeting of offers and optimization of sales strategy is the key focus area now. Another major driver to leverage AI is the compelling need to comply to the stringent regulations across different areas where the banks operate.
Finally, financial institutions are realizing the substantialcapability of enhancing marketing effectiveness and customer service through automation.
Surge in Use Cases
Let’s discuss some of the emerging AI applications and machine learning that have gained prominence recently. While banks, asset managers, and insurers have employed various pilots across the front, middle, and back offices, here are some of the key initiatives:
Credit and Insurance Underwriting: The lending and insurance department has rolled out machine learning to process the lending and insurance applications swiftly without incurring extra cost nor compromising on the risk assessment standards. The core message is machine intelligence definitely excels in accuracy, scale and speed as compared to human intelligence while sorting and analysis of colossal bulks of consumer data.
For example, banking and insurance sector has already started using self-learning algorithms to pittons of consumer data sets factoring age, job, marital status, credit history, and similar gradients in order to flag the risky profiles of individual applicants based on the generated information.
Fraud Detection: Previously, teams of finance experts used to follow a standard conventional checklist of risk factors and some complicated set of guidelines to detect fraud. However, ML-driven fraud discovery system proactively picks the irregularities and flags it for the security people to immediately investigate the matter in depth. The potential of intelligent algorithms to proactively forest all the possible fraud amongst a pile of infinite data sets is quite helpful in reducing false positives, wherein an anomaly is flagged and found out to be a false alarm.
Work flow Automation: Bank managers and insurers have already employed natural language process (NLP) to seamlessly automate some of their processes to rationally cap their operational costs and also enhance the customer satisfaction. Chatbots replacing the humans on the customer service interface has gained quite some traction.
‘Yes Robot’ of Yes Bank is a Personal Banking Assistant that helps customer conveniently check balance, recent transactions, send and transfer money, recharge and pay the phone bills, check loan eligibility and many more services. It also helps locate nearest bank ATMs & branches. It can be accessed through the applications and interfaces such as the Facebook Messenger.
ICICI Bank’s ‘software robotics’ is a kind of software that automates, regulates  and performs tasks of high-density and volume that needs to be  carried out over multiple applications while also increasing productivity. It uses facial and voice recognition, NLP, Machine Learning and bots to automate more than 200 business processes. It uses algorithms to sort processes & connects internal applications to external ones such as Aadhar or PAN card verification for KYC compliance. A sequential decision-making method is then used to sort the processes.
Asset Management: Termed as “robo-advisors,” like the Betterment and Wealthfront, are globally providing algorithm-based, automated financial planning solutions to their clients, serving them in developing investment portfolios that are aligned to their individual goals and risk tolerance.
Back home, FundsIndia.com has a robo-advisory service and looking for partnerships with financial giants. 15% of the company’s overall portfolio comprises robo-advisory services. Similarly, 5nance has an agreement with HDFC Mutual Fund for its robo-advisor.
Algorithmic Trading: Hedge funds and many other trading platforms in finance marketing, use complex AI algorithms to transact millions in the stock market every day, day after day. These systems, that are derived from machine learning and deep learning, facilitate “high-frequency trading” (HFT) while scrutinizing vast volumes of market factors in real time.
Conclusion
AI has umpteen applications in the finance and insurance environment, positioned to give the entire industry huge facelift in years to come via detection and analysis of brand sentiment; providing investment insights; making banking more efficient and less risky, and identifying fraud proactively.
As the rise in technological innovation is breaking the glass ceiling, the finance and insurance sector imperative must be to prioritize their goals and establish long-term strategies. We have been witnessingclear indicators of confident maturity among CIOs as they weigh their investments.
Liked our blog, pleaseleave your comments. We value your thoughts to make this section more interesting.
References
Nagasundaram, M. (2018). Retrieved from www.hcltech.com: https://www.hcltech.com/blogs/rise-machines
PWC. (2018). Retrieved from www.pwc.com: https://www.pwc.com/us/en/industries/financial-services/research-institute/top-issues/artificial-intelligence.html
Sigmoidal. (2018). Retrieved from https://sigmoidal.io: https://sigmoidal.io/real-applications-of-ai-in-finance

0
Posted on Author admin Categories IT infrastructure Management

Network Performance Management: 2018 Themes to watch out

A recent study states some interesting facts on network management trends, here are some of them that you should be knowing before making an informed and strategic decisions for network continuity.
We are in an always-on and ever evolving thrilling era of IT. There is a myriad of technologies that have been changing the way networks are built and accessed, how the data is transmitted and stored. Artificial Intelligence and machine learning, Cloud computing, IoT and many others provide unique opportunities for enterprises to digitally transform their business operations.
As diversified these technologies are, what unifies them is their dependence on a robust network functionality, otherwise known as ‘network continuity’. The core component to achieve that is visibility.
It is a known fact that all the latest advancements have led to driving networking best practices. With a bulk of business objectives and other critical activities heavily relying on IT, network performance is truly a matter of do or die for most of organizations.  Therefore, it has become a business imperative that enterprises keep a firm grasp on the latest trends to ensure they make informed and strategic network management decisions.
To follow these trends closely, EMA (Enterprise Management Associates) released 2018 edition of their bi-annual management study. Whether it is vast outcome from cloud services and networking toolset challenges, convergence of network operations (NetOps) and IT security. The reports sheds light on various captivating themes that have evolved the network management processes and the resulting impact it has had on businesses.
New Initiative influencing network management priorities
Since past few years server virtualization was the key driving factor for network decision making by a huge margin, almost 50% of the IT companies cited that to be their 2016 top initiative. Merely virtualization will not work as we tread through 2018.
Software defined data centers (SDDCs), public cloud or infrastructure as a service (IaaS), and private cloud initiatives are now the most influential drivers behind network management decision-making.
Enterprises now require all-inclusive and deeper network performance through a host of new efficiencies leveraging the latest technology of cloud and SDDCs, managing the complexities of network processes.
With growing network complexities, complexity of understanding and resolving performance issues also has to keep pace.  It is possible only through visibility into every segment of network transactions as they traverse physical networks, virtualized environments, and the cloud. Only then we can effectively identify, troubleshoot and resolve network issues, regardless of their origin.
Cloud services are flooding enterprise networks
With growing adoption of cloud, the entailing impact on the networks seem to be an important driving force behind IT decision-makers. As per EMA survey, 60% of participants suggested external public cloud traffic to be the workload presence on their network, of which 50% of their traffic volume can be traced back to their public cloud. Network performance monitoring and management can be a daunting task during cloud saturation, specifically in the absence of necessary visibility.
In practice, only 15% of network managers stated that they could oversee cloud networking with current solutions. The reason being maximum management solutions are not built to do that. Over 60% opine that they need to attain some new monitoring and troubleshooting tools for cloud services, while 14% are presently still on the hunt for the right solution.
Accurate cloud visibility solutions depend largely on what cloud is being used for. Software-as-a-Service (SaaS) functions need monitoring their service levels from the outside looking in, on the other hand Infrastructure-as-a-Service (IaaS) platforms may be greatly monitored in conjunction with the applications they are running.
The dawn of cloud services has stimulated an undisputable necessity for better insight into performance across hybrid environments.
Patchwork management solutions plague NetOps
The top most challenge in 2018 for NetOps is fragmented management solutions. 75% of IT businesses are using over ten active tools to monitor and troubleshoot their networks.
Unsurprisingly, those NetOps that are dependent on crowded roster of solutions in all probabilities will fail to detect network issues, as a result surely suffer higher volumes of network service outages annually.
Visibility is a hurdle to network operations using larger toolsets. What that means for network teams?
Often using too many specialized management solutions leads to a chaos causing miss out on an in-depth network insights compared to those network teams who easily manage only a few yet feature-rich solutions. Irrespective of the bulk of budget you allocate or use countless resources, it is highly impractical and insurmountable task to train and effectively enable your in-house network operations team to manage wide range of tools.
As opposed to that issue, users wrongly opt for their personal likes to tools, unaware of their loss of visibility and functionality. Addressing “tool sprawl” by consolidating scattered network solutions is effective and economical.
NetOps and IT Security are working in sync
Gone are the days when NetOps and IT security teams worked in silos, collaboration between the two is quite common now.
Going by the trends, 40% of EMA survey participants answered that they are completely converged with IT security, whereas 35% of enterprises have initiated the task of using security risk reduction as a yardstick to calibrate their network management achievements. And many of the network supervisors recognized their network performance monitoring and advanced network analytics as the top operations priority requiring integration with security processes.
What is the driving force behind the upward trend in collaboration between NetOps and IT security?
Enterprises have understood that these functions are more effective while in co-ordination than in isolation. The level of teamwork between NetOps and IT security will continue to grow transversely, with a shared intention of building robust network security.
Data source continues to change
In current times, the most widespread data sources in use for sustained network availability and performance monitoring are network test traffic, management system APIs and packet inspection. Additionally, the maximum prevalent data sources in use for network troubleshooting tasks include management system APIs and packet inspection.
It is in the amalgamation of insights from numerous data sources that the future of network management is possible. Firstly, the important step is to coordinate across data sources. But higher levels of coordination yield a great deal of further insight. Imagine the power when the broadest, most efficient view triggers greater attention to a specific area, and that greater attention yields specific insights that are examined in-depth. In practice, NetFlow is enabled to exactly pinpoint where the problem is stirring, deeper flow analytics can identify the problem area, and network packets can discover the actual problem source.
Outsourcing of network management
As per a report, 58% of businesses are outsourcing few or all aspects of network management, which has been constantly rising since 2014. This shift is a clear indicator of where a major chunk of IT market is headed towards – support from managed service providers (MSPs).
Presently, enterprises are outsourcing their managed services lot from WLAN networking and support, 24×7 network health monitoring, and data center monitoring, up to direct infrastructure management and configuration.
The challenge lies in whether to outsource, what to outsource, and before all that, how to ensure the transition and subsequent operation are successful, it has never been more critical for internal network managers – as well as external MSP partners – to have access to in-depth data on all network performance trends and anomalies.
Conclusion
So, what really connects these network management themes in 2018? What is the common thread?
The answer is actionable visibility – amidst all the evolving and ever-changing trends, network continuity completely relies on your enterprise or IT partner ability to not only achieve insight into what is happening on your network, but the agility and efficiency that you can do something about it.
As years go by there will be latest trends surfacing, but the consistency of IT partners should be maintained in keeping themselves educated and learn to seamlessly implement the latest trends and tools to gain network continuity through actionable visibility in their network performance!
References
Morville, P (2016): https://searchnetworking.techtarget.com/feature/Network-security-vs-network-performance-the-line-is-blurring
Zulch, L. (2018): https://www.networkworld.com/article/3285658/network-management/6-key-themes-shaping-the-future-of-network-performance-management.html

0
Posted on Author admin Categories Cloud, Healthcare

Healthcare Sector and Cloud Computing: Transforming to Serve better

Technology is continuously developing, especially in the highly competitive healthcare industry. Going by historic data, healthcare industry has been one of the slowest adopter of technology; and surely for valid and obvious reasons to stay vigilant and conventional in approach.
Enter cloud computing!
Cloud computing is altering the way healthcare providers deliver quality at affordable services.
The global healthcare cloud computing market is forecasted to become $9.48 billion by 2020 from $3.73 billion in 2015 — growing at CAGR of 20.5%. The market will be dominated by North America, with Europe and Asia to follow. The growth will however, pressurize healthcare system infrastructure to maintain and improve access to quality care without overburdening the costs.
Healthcare providers have no choice but to embrace the cloud in some form. This transition is being driven by two forces: the business imperative to cut costs and facilitate better quality of care.
Let us discuss some of the benefits that cloud computing has been able to bring about in healthcare sector.
Access to Healthcare
Getting access to proper healthcare in remote areas is a greater challenge, mainly when patients have a busy schedule to chase. In the midst of the lifestyle upheaval, telehealth and virtual care solutions are gaining impetus with laws also being modified to accelerate adoption and reach out to the needy at large.
Medication adherence
Patients falter in following prescriptions as advised that many times leads in their re-admissions, costing a huge sum on healthcare insurance units and indirectly on the government.  Automated messaging to keep a log of medicine refills before they expire or when to consume is expanding rapidly to avoid such burdens.
Drug theft and counterfeiting
Theft, counterfeiting, selling expired medicine are some of the problems which can be controlled by monitoring the supply chain. This has opened a vast market for solutions which monitor and log supply chain procedures in real-time and report suspicious actions.
Resource Incompetence
Mounting expenses of healthcare is debated the most amongst policy makers and yet no real solution has been employed to date effectively. One of the key factors that adds to the cost of healthcare is inadequacy of resources like medical staff, equipment and easy access to patient resource pool for clinical studies.
With the use of artificial intelligence in the healthcare environment, healthcare experts’ capabilities can be expanded since data can be amplified with smart machine-based analytics for doctors to appraise. For clinical trials and scientific studies, a social network-based approach can be used to gain accessibility to the patient pool.
Personal data privacy
Every healthcare organization that maintains and manages their own medical records witnesses a nightmare on the data security and compliance front, let alone that it adds significant cost for them to maintain their own IT infrastructure and be liable for all the data directly.
Cloud-based solutions provide access to state-of-the-art security technologies, thereby minimize the individual liabilities on each of the healthcare organization.
Uniform medical records
Each hospital or care provider who have in-house customized Electronic Health Record (EHR) system is not in favor of the consumer change. In addition to overburden the cost to the healthcare system due to the hassle of maintaining a different system for each hospital, it also makes more painful for patients to change their practitioners. What that means for patients is, they have to remain confined to a certain care provider and may not always get the finest care which could be the case if they had easy access to their EHR in a typical uniform format.
Conclusion                                                                         
As digital transformation spreads across the healthcare industry, improvements in connectivity, security, and cloud services technologies are allowing the healthcare ecosystem to solve numerous major challenges that the sector is facing — leveraging health clouds to address key issues.
Cloud computing and healthcare industry are a perfect match – together they can give fantastic health services and reach out to the otherwise inaccessible patients. With recent advancements in cloud computing the healthcare ecosystem is positioned to make the most of networked applications, as a result create and deploy better healthcare solutions.
References
Gupta, V. (2018) Cloud Computing in Healthcare. Retrieved from https://www.cisco.com/c/en_in/about/knowledge-network/cloud-computing.html
Patel, A. (2018, Jan) 6 Ways Cloud Computing Is Transforming Healthcare Systems. Retrieved from https://www.eetimes.com/author.asp?section_id=36&doc_id=1332854

0
Posted on Author admin Categories IoT

Digital Twin 2018: Technology and Simulation

Digital twin is the virtual depiction or replica of any physical object or system throughout the lifecycle of it by the inputs from real-time operational data and various sources to bring about better knowledge about the asset for informed decision making. A physical object can vary anything as small as a ball bearing that requires electrical, mechanical and software precision and seamless inter-operation.
Digital twin has moved past the manufacturing sector and has merged with Internet of Things, Artificial Intelligence and data analytics. As more and more complex objects are being connected, the process enables to produce more data to develop a digital replica that further enables scientists and engineers to gather in-depth information to optimize peak efficiency of the object before physically installing it or implementing it by developing scenarios of future possible breakdowns with probable solutions to it.
Digital Twin Application
Mostly used in manufacturing setup, digital twin is quite an advantage in the energy, transportation and construction sectors too. Large and complex objects like aircraft engines, turbines and trains could be digitally designed and tested before being produced. These digital twins are also helpful in maintenance operations, engineers or technicians can seek the digital twin to test before possible fix or upgrade a part of a specific equipment before fixing it on its physical twin.
Creating digital twin with umpteen data requires highly sorted skill sets such as machine learning, AI, predictive analytics and many more science capabilities.
Digital Twin and IoT
With IoT becoming universal, the device sensors used in IoT can be used by digital twin to include small and less complex objects, offering added advantage to companies.
An article by Dave McCarthy cites the reasons of having digital twin to deploy IoT, including the ability of digital twin to foresee various outcomes depending on the variable data. Similar scenario ‘run the simulation’ is often noticed in the Si-Fi movie, where a fictious scenario is depicted within a digital space.
Using added software and data analytics, digital twin can be leveraged to deploy IoT devices for maximum efficiency. In addition, designers can leverage digital twin to ascertain fixing each part into exact device and how to operate before physically deploying. The more a digital twin can replicate a physical object, the higher is the likelihood of replicating the efficiency levels.
According to Dean Hamilton, digital twin and IoT together can revolutionize the manufacturing world. “The more highly instrumented a device is, the more accurately its digital twin will represent its actual historical performance, leading to better analysis and simulation of its future performance,” Hamilton writes.
Problems with Digital Twin
The general approach of digital twin is to demonstrate an insight of the objects or products at operational stage, without highlighting other features or making any comparative study of the product when designed, built and other dynamics of it. It involves a platform creation to line up the characterization of virtual object to translate the complete operational data into the digital twin which enables to comprehensively understand the product performance in comparison to its design intend.
Though digital twin is a virtual image of the product asset in usage in addition to certain predictive analysis and visual elaboration of lifecycle based on predictive algorithm, yet it’s not a complete replacement of ‘inclusive analytics’. Digital twin predictions cannot correlate and map back to product simulation, design modeling and overall lifespan predictive analysis destined at during the design and testing phases. Digital twin can only predict the requirements of parts replacement, it cannot predict a flaw in specific lifecycle of the asset.
Options for Analytics
AI vendors adopt a collective approach of collating and sending all the viable data to create a composite virtual image of a physical object. A virtually created giant asset such as a car has the capability of sending 25GB or more data per hour back to the cloud. The need is creation of algorithm analytics for running the data that is constantly created at the edge of the enterprise network. Analytics or cognitive automation is a mandate to recognize and deliver data to the cloud pro-actively before the networks hit a complete gridlock, that Immersive Analytics vendors often fail to acknowledge nor address.
 Digital twins can greatly augment an enterprise’s ability to make proactive, data-driven decisions, enhancing efficiency and evading potential problems. However, it can be a huge challenge for companies to create a digital twin if they would like to try this all at once. Best option is to start in one area, deliver value and continue to develop.
 
Reference
https://www.zdnet.com/article/the-rise-of-the-digital-twin-why-the-enterprise-needs-to-take-notice/
https://www2.deloitte.com/content/dam/Deloitte/cn/Documents/cip/deloitte-cn-cip-industry-4-0-digital-twin-technology-en-171215.pdf
https://internetofbusiness.com/half-of-businesses-with-iot-projects-planning-to-use-digita-twin/

0
1 … 11 12 13 14 15 16