• Calender

    March 2013
    M T W T F S S
    « Feb   Apr »
     123
    45678910
    11121314151617
    18192021222324
    25262728293031
  • Contact

    Send us press releases, events and new product information for publication..

    Email: nchuppala@ outlook.com

Best Sellers on Predictive Analytics

Predictive analytics is red hot. Why? What organization couldn’t benefit from making better decisions?

Just ask the Obama campaign, which used sophisticated uplift modeling to target and influence swing voters. Or telecom firms that use predictive analytics to help prevent customer churn. Or police departments that use it to reduce crime. The list goes on and on and on. Virtually every organization could benefit from predictive analytics. Don’t confuse traditional business intelligence (BI) with predictive analytics. BI is about reports, dashboards, and advanced visualizations (which are still essential to every organization). Predictive is different. Predictive analytics uses machine learning algorithms on large and small data sets alike to predict outcomes.

But predictive is not about absolutes; it doesn’t guarantee an outcome. Rather, it’s about probabilities. For example, there is a 76% chance that this person will click on this display ad. Or there is a 63% chance that this customer will buy at a certain price. Or there is an 89% chance that this part will fail. Good stuff, but it’s hard to understand and harder to do. It’s worth it, though: Organizations that employ predictive analytics can dramatically reduce risk, disrupt competitors, and save tons of dough. Many are doing it now. More want to.

Few understand the what, why, and how of predictive analytics. Here’s a short, ordered reading list designed to get you up to speed super-fast:

The Signal And The Noise: Why So Many Predictions Fail — but Some Don’t” by Nate Silver. Nate Silver built an innovative system for predicting baseball performance and predicted the 2008 and 2012 elections within a hair’s breadth — all by the time he was thirty. The New York Times now publishes FiveThirtyEight.com, where Silver is one of the nation’s most influential political forecasters. Why read: This book will simultaneously inspire your “predictive” imagination and ground you in the realities of predictive analytics.

Predictive Analytics: The Power To Predict Who Will Click, Buy, Lie, or Die” by Eric Siegel. Former Columbia University professor and Predictive Analytics World founder Eric Siegel has written a very accessible book on how predictive analytics works. It’s chock-full of dozens of real-world examples, such as how Chase Bank predicted mortgage risk (before the recession), IBM Watson won Jeopardy!, and Hewlett-Packard predicted employee flight risk. Why read: Predictive Analytics is a perfectly paced explanation of how predictive analytics works and a repository of dozens of real examples across many use cases.

Uncontrolled: The Surprising Payoff Of of Trial-and-Error For Business, Politics, and Society” by Jim Manzi.  Predictive analytics is not magic — it’s science. Jim Manzi reminds us of the power of the scientific method and reveals the shocking truth that many huge societal and business decisions are made based on  misinterpreted data and statistics. Why read: Controlled experimentation amplifies the results of predictive analytics and helps avert the risk of inaccurate predictive models.

Data Mining: Practical Machine Learning Tools and Techniques” by Ian H. Witten, Eibe Frank, and Mark Hall. If you write code or have a computer science background, you’ll probably want to know the gory details of how predictive analytics works. Why read: Data Mining provides a thorough grounding in machine learning concepts as well as practical advice on applying machine learning tools and techniques in real-world data mining situations. 

The Forrester Wave: Big Data Predictive Analytics Solutions, Q1 2013 (Forrester client access only). Forrester assessed the state of the big data predictive analytics market. To see how the vendors stack up against each other, we evaluated the strengths and weaknesses of top big data predictive analytics solutions vendors, including Angoss Software, IBM, KXEN, Oracle, Revolution Analytics, Salford Systems, SAP, SAS, StatSoft, and Tibco Software. Forrester expects the market for big data predictive analytics solutions to be vibrant, highly competitive, and flush with new entrants over the next three years. Why read: These vendor solutions can jump-start your predictive analytics program.

Happy reading, and please recommend some other good predictive analytics reads.

This blog originally appeared at Forrester Research.

 

 

The Culture Change Needed for an Analytics Supply Chain

The explosion of data collection promises so much. Where once companies were only able to collect structured data from limited sources; now they can mine everything from social media to call center transcripts, with more and more big data tools available to process a far wider range of information. So, those who collect the most data are best positioned to gain a competitive advantage, right?

 

Not so fast. When companies try to use their newfound wealth of data for a specific purpose (e.g., to decide whether to launch a new product), they invariably discover some glaring gaps in the information they require. Why are companies still finding themselves short of the right data to answer key business questions? The answer is that they didn’t design their data collection processes with those exact questions in mind.

Getting past this challenge requires a shift in mindset. Data collection isn’t a goal in its own right, but a means to an end. And getting to that end – where the company is able to use data to answer more of its questions – requires applications that are designed accordingly. The applications still need to meet users’ needs, but their data collection features need to be tailored to produce the answers to the questions the business is actually asking (or expecting to ask).

To be sure that the right data emerges at the end of the collection process, companies need to address this problem as a supply chain issue.

By quantifying every aspect of their businesses’ actions, interactions and processes, firms should be able to decide which questions they need to answer in order to be smarter. The necessary data required for that will then be clearer, so the challenge becomes how to collect it or even how to create it.

And help is available for that. Software vendors offer tools that enable companies to extract data more easily from their packaged software applications. New instruments can be added to existing applications to capture additional data. For example, Netflix tracks how its customers interact with its streaming movies, studying metrics such as when viewers pause the movie and which scenes are watched repeatedly. This data has helped the company improve its recommendation engine.

Sensor technology enables companies to create and collect information from physical environments. Look at UPS: The company’s in-vehicle sensors have tracked movements of its trucks so precisely that the company determined that making left turns slows deliveries and increases fuel costs. By redesigning routes to minimize left-hand turns, it has saved 9 million gallons of gas per year and improved customer service by predicting more accurate delivery times.

The process of quantification has evolved quickest when it comes to consumer behavior, where opportunities for data collection have never been richer. For instance, there are social media applications where consumers provide their own data, like the online reviews found at Yelp. Another example is in the area of personal fitness, where applications, such as Nike+ FuelBand and Fitbit, collect data on everything from consumers’ calorie burning to their sleep patterns and diet.

Other data sources for analysis include unstructured data, such as email and tweets, as well as audio files from call center conversations. And companies should not neglect the data collected by others with which they work. Service providers, social networks and even software providers running packages in businesses’ data centers may all be useful sources of information.

But only those companies that know what kind of questions they want answered can judge which of these data sources will truly help them. So the supply chain continues; having identified where the right data will come from, the next step is to organize and manipulate it for analysis.

Companies that expressly design the supply chain with specific analytics in mind will find it more straightforward to bring the data together to answer those questions because there are far fewer information gaps to contend with. The ideal is a virtuous feedback loop in which companies collect, analyze and respond to data in an increasingly agile manner; the questions are revisited and revised on a regular basis as business conditions change and the company’s strategy evolves.

However, getting to that stage will require IT to work much more closely with business leaders. These leaders are responsible for deciding which questions are most important to the business, and surprisingly few executives have processes in place for documenting those questions and sharing them. By setting up such processes, IT and business can work together to explore all the possible sources of data that might provide answers. They will also need to establish how those answers will be distributed within the business so that they are actionable.

This is a cultural shift for most businesses. In order to design effectively for analytics, the roles of the IT organization, which provides the data, and the business, which consumes it, need to be much more blurred. The aim is an enterprise culture where IT and business work together closely to deliver insights that produce a real competitive edge.

Some companies have addressed this objective by creating a new executive role: a chief data officer who can champion the collection, prioritization, distribution and analysis of data. While this push from the top has been effective for many organizations, the wider goal is a culture where all employees are more data-aware. Data technology has evolved but people still must decide what data is useful and how to harvest it.

The prize is a valuable one. Already, data is available from more diverse sources than ever imagined for businesses to analyze and apply. But businesses need the right data, and companies that design their applications, processes and products to deliver data-driven insights will be the ones that succeed and prosper.

(Author’s note: To read more about “design for analytics” and other trends IT and business leaders need to prepare for, download Accenture’s newly releasedTechnology Vision 2013.)

Narendra Mulani is the managing director of Accenture Analytics. Leading an integrated community of more than 15,000 management consulting, technology and outsourcing professionals who serve clients around the globe, he is responsible for driving Accenture’s strategic agenda for growth across business analytics. Mulani is a member of Accenture’s Global Leadership team. He graduated from Bombay University in 1978 with a Bachelor of Commerce. He received an MBA in finance in 1982 and a Ph.D. in multivariate statistics in 1985, both from the University of Massachusetts. Prior to joining Accenture, Mulani ran his own consulting company.

 

 

The Performance Dashboard: The New Face of Business Intelligence

by Shaku Atre

To stay competitive, organizations need business intelligence – accurate, actionable information about key business metrics — more than ever.

A performance dashboard, the user interface ,providing summaries and reports of the most critical information which is called business intelligence delivers the most business value when used to manage progress towards operational, tactical and then strategic goals based on Key Performance Indicators (KPIs).

Creating an effective performance management dashboard with the proper mix of pictures, photos, diagrams, graphs and charts requires business, technical and political skills. It is an ongoing process of understanding the changing needs of the users that will be served, the information that is most critical to them and the presentation methods that are most effective.

Organizations need dashboards because everyone from senior management to “front-line” employees are inundated with data in the form of reports, memos, emails and phone calls, among other electronic and non-electronic formats. This data needs to be synthesized into the critical information that will enable them to make decisions, take actions or delegate work in a timely way. While many “dashboards” are only digital versions of paper reports, an effective dashboard dynamically combines information in new ways to measure performance (where appropriate in real-time, such as in stock trading or the condition of Intensive Care Unit patients) and enable fast, appropriate action by decision makers , whether in management or the “trenches” .

What Makes an Effective Dashboard?

Presenting the right information, to the right people, at the right time, in the right presentation format and at the right cost make a performance dashboard successful. If a dashboard is not easy, or even enjoyable, to work with, users will abandon it. But if it is only pleasing to look at and doesn’t give users the information they need to make timely decisions they will also abandon it. Decision-makers must find the dashboard easier than relying on less accurate ad-hoc reports or the corporate grapevine. Ideally, a dashboard should encourage not only its use but interactivity. When a decision maker creates new views of data or share ideas with other dashboard users, they help assure that problems are detected and resolved in the most appropriate timeframe at the right cost.

An effective dashboard allows users to change the type of information they see, or how it is displayed, without help from IT as business challenges evolve. For example, during a downturn a user might need to track KPIs focused on cost-cutting. When the economy improves, those KPIs might shift to market share or increased sales. With the acquisition of another company, a user might suddenly need to track the performance of new product lines, or merged product lines, or sales in a new geography. Most importantly, an effective performance dashboard must clearly and quickly alert decision makers when they need to take action, such as if customer retention falls in a specific region or defect rates rise on a certain production line.

Finally, a performance dashboard must increase both the speed and accuracy with which managers make decisions. This requires that accurate data can be quickly found and extracted from operational data stores.

Major Types of Dashboards

There are three major types of performance dashboards, each with different requirements for level of summary, analytic capabilities and user interfaces.

The first, and most numerous, type of dashboard is the operational dashboard. These are usually deployed at the departmental level to monitor the processes that generate products or services. They need to provide as much detailed data as possible, reducing the need for ‘drill downs” because they are used by employees in the ‘trenches” who don’t have time to analyze a situation and often have the experience to solve problems without further study. For example, if a delivery route is hit with an unexpected blizzard dispatchers have to redirect trucks instantly to keep deliveries on schedule. For a manufacturing firm, an operational dashboard might track products manufactured along with the number of defects, complaints or returns. For retail, it might track daily revenue or sales based on product placement to make the best use of valuable shelf space.

The second, and somewhat less numerous, type of dashboard is a tactical dashboard, which monitors the processes that support the organization’s strategic initiatives. A tactical dashboard falls between the strategic and operational dashboards in the level of detail at which the data is presented. If, for example, an organization wants to grow worldwide market share by 15 percentage points in a year, a tactical dashboard will track overall actual vs. target market share in various geographies, while the operational dashboard would track sales of specific products against their competitors at different times throughout the year. A tactical dashboard thus requires greater “drill down” capabilities than an operational dashboard because it presents information in a more summarized form. The interface tends to be more graphical than that of an operational dashboard, and very clearly tracks performance of the KPIs.

The third type, a strategic dashboard, measures progress towards enterprise-wide strategic goals. It might include non-financial KPIs such as customer satisfaction, customer attrition or market share (as well as comparisons such as this year’s fourth quarter sales vs. the same quarter in prior years) along with financial KPIs such as profitability and lifetime value per customer. Since a strategic dashboard presents highly summarized global trends for top management, it must provide greater abilities to drill into the underlying data. It also requires less frequent updates from supporting databases. Along with internal operational and tactical trends, it is also more likely to include data on global, external trends (such as the market share or projected profit margins of competitors, currency fluctuations or legal or regulatory trends where applicable for industries such as pharmaceuticals. It may also include information about external events such as political unrest that might have an impact on the strategic goals and require action by top management.)

Ongoing Effort

The relationship among the three types of dashboards can be represented as an upside-down, hierarchical tree structure with strategic dashboards at the top, major branches made up of tactical dashboards growing down from the strategic dashboard and smaller branches and leaves of operational dashboards hanging from the major branches.

Keeping a dashboard effective over time may require changing everything from the type of data presented, the level of detail at which it is presented, the presentation format or even the KPIs themselves if feedback from the operational and tactical dashboards shows that the KPIs set in the strategic dashboards are unrealistic.

Rule of Thumb: Identify early on which type of dashboards you need, because each has very different requirements for data, user interfaces and reporting. Your dashboard project will involve a learning curve as you continually iterate through three steps:

  • Step #1: Study the user requirements at all levels;
  • Step #2: Plan and implement a dashboard with user input, with prototypes if one can afford the expense
  • Step #3: Learn and improve

Summary

Dashboards put a new face on traditional business intelligence by providing timely actionable information in the form of easily understandable diagrams, pictures, charts and signals rather than the columns of numbers found in traditional reports. Their purpose is to identify whether the organization is meeting, exceeding or falling short of the key performance indicators that drive business success. The presentation must be simple, clear and easy to use, and the dashboard must be flexible enough to change as business needs evolve.

Shaku Atre è una speaker eccezionale che ha la reputazione di catturare l’attenzione dei partecipanti e di mantenere vivo l’interesse anche in presenza di argomenti complessi. E’ presidente di Atre Group, Inc. una società di consulenza, training e publishing nel settore della Business Intelligence. E’ stata Partner in Price Waterhouse e 14 anni in IBM. E’ una esperta rinomata nei settori del Database Management e del Data Warehousing. Ha tenuto seminari su questi temi in USA, Canada, Europa, Asia e Sud America. I suoi articoli sono frequentemente pubblicati in Computerworld, Information Week, Information Management, Tech Web e altre importanti pubblicazioni di computer. Ha scritto numerosi libri fra i quali ricordiamo il best seller “Database: Structured Techniques for Design, Performance and Management” pubblicato da John Wiley and Sons, che ha venduto più di 250.000 copie ed è stato adottato da molte importanti Università tra cui Harvard, Columbia, Cornell, MIT, New York University, Stanford and U.C. Berkeley. Altri libri “Information Center: Strategies and Case Studies”, “Database Management Systems”, “Distributed Databases, Cooperative Processing & Networking” ed inoltre “Atre’s Roadmap for Data Warehouse/Data Mart Implementation” pubblicato da Gartner Group. L’ultimo libro pubblicato è “Business Intelligence Roadmap: The Complete Project Lifecycle for Decision-Support Applications”.

This article is based on a two-day seminar Ms. Atre will present for Technology Transfer in Rome, Italy April 13-14th, 2011 providing a complete guide for planning, designing, implementing, using, and maintaining performance dashboards. Meant for business managers and IT specialists, it will walk attendees through detailed role-playing exercises and equip them to deploy dashboards when they return to their organizations.

EVENT : On Dashboards by Shaku Atre

Shaku AtreMarch 27, 2013
Rome calling!

How about joining me at the Rome event sponsored by Technology Transfer Institute on April 8-10,2013, on Dashboards and on April 11 , on what do you think, onBig Data? Our last year’s events in Rome were a huge success – as a result this year’s events are scheduled by popular demand! You don’t want to miss it – and don’t forget that Rome is gorgeous in April!

Want to learn how to create strategic, analytic and operational dashboards for your users? And want to know how to effectively implement Big Data ? Then these two seminars are for you. By the way these seminars will be held in an old , completely renovated, absolutely beautiful seminary.

Join Technology Transfer Institute in Rome  for these two  seminars that will explore the strategies and processes to take raw data and build a fully functional dashboard and will present my Ten Golden Rules of Big Data.

In addition to outlining the steps on how to create effective dashboards, you will receive hands-on training on how to design, implement, use, maintain, revise and enhance your dashboards.

Please download the brochures to see what you will be missing if you don’t join us in Rome.

» DASHBOARDS: Is it the new face of Business Intelligence? (April 8-10, 2013)
» Ten Golden Rules of Big Data (April 11, 2013)

 

Best Regards,

Shaku Atre
Atre Group, Inc.
shaku@atre.com
www.atre.com

Managing Teams & Giving Feedback – Project Management

Constructive feedback

NuoDB Now Available on Amazon AWS:

NuoDB on Amazon AWS Marketplace, accelerates your ability to develop, deploy and elastically scale apps in the cloud. And you can start developing your next cloud application on NuoDB on AWS at no cost.

Get Started With NuoDB & Amazon:

    1. Learn how to provision NuoDB on AWS using the Amazon Cloud Formation Script in our Tech Blog.

    1. Visit NuoDB on the Amazon AWS Marketplace and start your instance.  

  1. Check out the NuoDB documentation for further information on NuoDB.

As a NuoDB downloader, you can easily upgrade and migrate your cloud database apps to Amazon AWS. Simply startup an Amazon instance with NuoDB as a peer to your existing database, and just like that you’ve migrated your database and doubled transactional throughput. 

Tell us what you think! Join the conversation on TwitterFacebook and on LinkedIn.

Happy Scaling!
The NuoDB Team

P.S. The latest release of NuoDB is now available. Download NuoDB Starlings Release 1.0.2 today for SQL query performance enhancements, improvements to Amazon S3 stability and more

 

THE MASTER OF SCIENCE IN ANALYTICS

Image

 

http://analytics.ncsu.edu/reports/msa.pdf