SigmaWay Blog

Hadoop has grown to be one of the best large scale and batch oriented analytics tool, used by webscalers as well as enterprises. Hadoop was designed to integrate with and complement the existing business intelligence of any corporation. But, the issue with Hadoop is that the adoption rate is very slow with the data center administrators. So, most developers have been looking at possible alternatives. Today we will name a few worthy alternatives that you can look at which have a potential of replacing Hadoop in the years to come. They are:
• Disco
• Misco
• Cloud MapReduce
• Bashreduce
• Qizmt
• HTTPMR
• Skynet
• Sphere
• Riak
• Octopy
• MapReduce
• Filemap
• Plasma MapReduce
• Mapredus
• Mincemeat
• GPMR
• Elastic Phoenix
• Preregrine
• R3
• Ceph
• QFS
• Cloud-Crowd
• HPCC
• Condor
• Storm
• HaLoop
• MapRejuice
• GoCircuit
• Spark
• Stratosphere
• Gridgain
• MongoDB
• Mars
• Minceat
• Dato Core
• HPCC
• MapReduce Lite
• Gearman

For more information visit:
http://www.fromdev.com/2015/03/hadoop-alternatives.html?m=0

As people are the most important asset for any organization, hence talent management is a crucial part for a successful organization. It is critical for an organization to control its people by talent acquisition, employee engagement, performance management and leadership and development for better performance.
The 5 Step paths to People Analytics:
v Identify the Stakeholders – Whichever be the area that might have been chosen, it is essential to identify everyone involved in that function.
v Establish Goals and Objectives – Quantifiable targets should be determined so that it can be assessed whether the target is achieved or not or for tracking the progress.
v Do a “reality check” – A realistic goal for improvement is to be determined.
v Bridge the gap – This step makes sure that everything that is needed for the process is present like people, processes and technology.
v Prove success with Data – Analytics in this step highlights the successes or the areas of improvement in the function.
Read more at: https://icrunchdatanews.com/5-step-path-people-analytics/

Cancer is a disease that had killed many people in past decades. Few years back it was not possible to treat a patient suffering from cancer. Now doctors have treatment of cancer. But it could more easier with the help of big data. Let us see how big data make it easier. 

Doctors collect data from pre and post treatment of patients. Predictive models can be formed using the extensive data that can help doctors to analyze whether a treatment is success or failure for some set of patients. There are some advancements where AI is used in diagnosis and treatment of cancer. With the help of big data analytics doctors can see which drugs are most effective in treatment. Also running analytical models doctors can find which drugs are not necessary and significant. Only necessary drugs can be used to target the specific forms of cancer. The systematic and data driven view for looking at cancer has increased the potential for its prevention. Read more at:https://channels.theinnovationenterprise.com/articles/big-data-in-cancer-revisited

A wealth of past data is available in years of medical records of ICU patients, which can be used to measure risk levels in ICU. In case of predictable risks, checklists can be created of obvious risks like infections caused by catheters or tubes, delirium caused by overseadation etc. In case of less predictive risk, a system can be generated that gives a running risk report of the ICU on the doctor’s hand held devices. Medical centers have developed an application that pulls out data from the EHRs and creates a visual dashboard of the patient that helps doctors in real time monitoring, thereby personalising patient care. Read more at:  http://www.wsj.com/articles/hospital-icus-mine-big-data-in-push-for-better-outcomes-1435249003

Big data can provide contextual intelligence, which would work as a catalyst in supply chain operations. While legacy ERP (Efficient Resource planning) and SCM (Supply Chain Management) are designed for order, shipment and transactional data, big data can create knowledge sharing supplier networks, revolutionizing the supply chain. Delivery networks can be optimized using geoanalytics, which fixes one of the major challenges in this business. Companies that have adopted big data analytics say that there is improvement in customer service and demand forecasting and optimization of inventory and asset productivity. Read more at: http://www.forbes.com/sites/louiscolumbus/2015/07/13/ten-ways-big-data-is-revolutionizing-supply-chain-management/

Sentiment Analysis is a text mining technique used to analyze the sentiment (unstructured data representing opinions, emotions or attitudes) about a particular product or topic, employing machine learning technique. Organizations across the world are extracting insights from the social data to understand how public feels about the product at a particular point in time. Sentiments are first categorized into negative or positive and then analyzed using various software like Hadoop. It can also be performed on excel. To know more, read the article at: https://www.brandwatch.com/2015/01/understanding-sentiment-analysis/

Through Predictive analytics, customer behavior can be predicted to learn engaging with them and improve their experience. Mobile on the other hand is making marketers refocus their analytic efforts. Thus, predictive mobile analytics enable organizations discover consumer behavior by looking at data of how apps are used and finding their acting patterns. Using this information, marketers can focus on their marketing and advertising initiatives, looking at customer engagement during a promotion in real-time to facilitate greater targeting. Through predictive analytics, one can find the targeted customers. Digital tags enable to create a digital dossier where browsing history can link to a particular unidentified individual helping maintain their privacy. Thus can establish repeat behaviors that lead either to a purchase, or a rejection. Read more about it at:  https://channels.theinnovationenterprise.com/articles/why-you-must-invest-in-mobile-analytics

To stay ahead of new disruptive competitors, banks must understand the value of the data produced from daily transactions across email, mobile and online channels by digitally-led customers and use them to build on their strengths. High street banks and private financial service organizations give customers potential to develop new initiatives with cross-marketing events, loyalty programs. Bank must adopt a mobile-first strategy which engages with customers to maximize longevity. They must deploy mission-critical analytics tools with access to real-time data. Also, more flexible and intelligent OS harnessing the use of big data must be used. By using sophisticated analytics features, banks’ risk management departments can access information on customer-purchasing behavior enabling them to make immediate adjustments to individual customer credit limits or lending rights. Read more about it at: https://channels.theinnovationenterprise.com/articles/7749-how-is-data-remodelling-the-fs-ecosystem 

A recent survey on customer retention revealed that insures lose as much as 20% of their customers each year. For the insurance agency this corresponds to approximately £500,000 to £1 million in lost premiums for every 1,000 policies. Customer retention is thus an important issue facing the industry. Three key points can help insurance companies to eliminate customer churn.
• Firstly, they need to properly understand their customers, which mean determining overall market segmentation to understand cross sell and up sell opportunities at individual policyholder level.
• Secondly, insurance companies must be able to provide tailor made policies for each individual and also have a unique selling point in comparison to their rivals.
• Lastly, companies need to realize that they have to listen to their customer’s preferences rather than the customer having to choose from the list of policies the insurer has to offer.
Should the above be kept in mind in a company’s planning and operations customer churn can certainly be minimized if not completely eliminated.

For more information visit:
http://blogs.sap.com/banking/2015/06/19/preventing-customer-churn-with-better-data-analytics/

A loyalty program is less about points and rewards and while these perks might attract consumers, they don’t suggest a sense of loyalty. In this age of data, the focus of these programs should be collection of useful data to help maintain good relations which benefit both consumers and the brand. However a business that trusts consumer-provided data is a business making decisions based on what is untrustworthy, random information. From observation of customer activities and external data sources, advanced analytics can create a profile of the customer for precise segmentation of the customer base. Business analysts and data scientists agree that expanding the data for any given model will typically produce dramatic improvements in analysis. That’s why many organizations have turned to a Big Data solution. Read more at: http://www.smartdatacollective.com/davemendle/323701/redefining-loyalty-programs-big-data-hadoop

Up to 16 Exabytes of RAM can be supported by a 64-bit system. Machines with 128GB RAM or more are becoming common with this era of Cloud Computing and Big Data. The data sets for Big Data are getting too large for even heavily loaded machines with memory despite the best efforts, don't fit into the RAM even after clustering in some cases. Researchers at MIT created a cluster called BlueDBM using Solid-State Drives (SSDs) to get rid of the memory problem. They also moved some of the computational power off the servers and onto chips. By pre-processing known parts of the data onto the flash drives prior to passing it back to the servers, the chips made distributed computation much more efficient than before. They thus got rid of the overhead of running an operating system. Read more at: http://www.itworld.com/article/2947839/big-data/mit-comes-up-with-a-no-memory-solution-for-big-data.html

Uncontrolled urbanization is posing as a great threat as it puts immense pressure on the resources available. Some of the issues that will be faced in this scenario are clean water, clean air, power, waste management and living space.
Big Data and Internet of Things (IoT) are some of the technological forces which are pushing towards smart cities. With the help of Big Data a range of applications are being made available with the help of which smart cities can be made more livable. Applications like smart street lights are already being used in smart cities like Birmingham. Street lights are attached with sensors which monitor footfall and noise levels and based on this data, the light turns on or off. Cities like Bristol, England are installing an infrastructural network that will support the data generated through Big Data and IoT.
These examples are minor steps in building of Smart cities which can sustain growing population without hampering the quality of life. Read more at: https://icrunchdatanews.com/big-data-building-smart-cities/

There is no doubt about the revolution that big data has brought to the way business is done. But, most of the talk has been around the structured data. It has been increasingly becoming clear that the potential of big data can be truly understood if we take up the challenge of harvesting unstructured data. Jonathan Buckley, senior vice president of marketing at Qubole, in an article in Smartdatacollective emphasizes that if businesses want to remain relevant and profitable then it’s the right time to turn their attention to text analytics. The most important advantage that text analytics have is that it provides with a much larger sample of customer sentiment and extract data which is otherwise not quantifiable. But all of this boils down to having the right technology. For more on this follow the link http://www.smartdatacollective.com/jonathanbuckley/329383/text-analytics-next-frontier-big-data

In today's world most of the jobs which were performed by humans have been replaced by robots and computers leading to increase in unemployment. For example: robot waiters, robot doctors and self-driving cars. A company in China is installing 30,000 robots every year that has snatched the work that was earlier done by human beings. Today robots are becoming cheaper so there is possibility that they might take over the jobs of human beings even in low cost companies. The main cause of this threat is the invention of smart machines. There is inverse relation between smarter computers and employment. Every year computers are becoming smarter, doubling their processing power and memories.  Machines that have Artificial Intelligent (AI) and smart computers that can act own their own have adversely affected humans, leaving them unemployed. Also high skills are required to operate these smart computers and not all humans got the same. According to some economists, AI should be controlled otherwise the results could be disastrous. Read more at:http://www.smartdatacollective.com/bernardmarr/330436/ai-biggest-threats-your-job

Permission marketing is now trending. Consumers don’t like to get interrupted often and so they take actions to block online ads, promotion emails. This new era of permission marketing can be broken down into “anticipated”, “personal” and “relevant”. A recently concluded study shows customers are willing to share information with trusted brands. Data privacy acts now protects consumers from personal data collection through websites and other online sources. Cookie law is such an example. Customer data can be explicit, implicit or structured by analytical tools. Thus data security is of utmost importance and it is the responsibility of organizations. Consumers are always looking for their needs to be met by companies and this gives the companies the chance to make the best use of data collected providing value to the organization. Read more at: http://www.cmswire.com/digital-marketing/tracking-customer-data-you-better-provide-value/

There are a lot of open source analytics tools available that every user can easily find on websites. But to make appropriate use of such tools each company requires equally compatible skills. So companies are required to invest heavily to develop such skills and also on collecting data. It has been found that companies are spending on an average of $7.4m on data initiatives in 2015. Free platforms also took huge investment such as Hadoop, making more efficient systems. The increasing data size led to increase in expenditure, and is likely to increase in future. Upgrade is required in currently installed systems to process huge data in less time. On the other hand huge investment is required to meet the heavy salaries of data scientists, which is approximately $118,000m. As the supply of qualified data scientists increase this amount will fall, which decreases the overall amount of employee compensation. There is a strong evidence that spending on data initiatives has been increasing at a surprising rate and is likely to continue for next 2 years which ensures a boom for the big data companies.  Read more at:https://channels.theinnovationenterprise.com/articles/big-data-spend-is-increasing

Data science is the most dynamic and multifaceted industry providing useful and interesting insights with the data. Recently due to more than exponential growth of data and digital revolution, it is becoming increasingly difficult for organizations to use this data efficiently. Sifting data has become cumbersome and getting to the right data has become a top priority. Thus most organizations are formulating strategies to use this data wisely without any hassles. From operational procedures to recruiting employees, right data can direct to right ways of doing it. Thus CEOs are showing more interest in this digitized environment. In short an overhaul in usage of data driven techniques are taking shape to have a leading edge. To  read more: http://www.predictiveanalyticsworld.com/patimes/plotting-your-data-science-strategy-0618152/

Data science with number of practical uses is becoming an indispensable tool for all businesses. Its pace has reached an exponential level in recent times due to many advancements. But many experts are now shifting the gears into next level, predictive analytics, to save the future of this invaluable science, but getting this accepted industry wide is going to be a rough ride. Getting a smooth change from data science to predictive analytics will need an industry wide trust. It also needs the barrier of welcoming the new entrant eliminated by showing the effectiveness of this new technology which will create a helping environment for employees and employers alike. But to many hardcore fans of data science it is tough time leaving it to accept a new technology, only the returns and competitive edge this provides will make the shift more pleasant. Read more at: 

https://icrunchdatanews.com/3-keys-smooth-migration-data-science-predictive-analytics/

The huge demand of predictive analytics has led to various trainings been sprung up to impart or at least promising to impart the kind of skills required for developing an analytics model. But more often than not it forgets the primary ingredient. Geert Verstraeten, managing Partner and professional trainer of Python Predictions, feels that understanding the main project phases is the most important thing. He explains these phases with an interesting analogy- developing analytics model is as simple as making soup in a soup bar. Here are 5 phases of model building: 

• Taking the order- Project Definition 

• Mise en place- Data Preparation 

• Cooking the soup- Model building 

• Tasting the soup- Model Validation

Dubai Airports CEO, Paul Griffiths, is revolutionizing the airport experience and working. The Dubai Airport is handling 120 million annual passengers. This causes a need for something preventing the degradation of customer experience. Using in-dept. analytics to assign gates in a dynamic way is the first step. The system can assign gates in a far more efficient manner – such that if two flights have a significant numbers of passengers common between them, they will dynamically be assigned gate positions as close as possible. A new initiative allows retail stores in the airport to scan passengers' boarding passes and provide current updated about their departure gate, time, and the estimated time to move from the customer's current position to the gate. There are thus using data to improve the efficiency and reduce the cost structure of air industry operations. Read more at: http://www.networkworld.com/article/2944535/big-data-business-intelligence/big-data-enables-an-entirely-new-design-of-airport.html