text.skipToContent text.skipToNavigation

Understanding the Predictive Analytics Lifecycle von Cordoba, Alberto (eBook)

  • Erscheinungsdatum: 30.07.2014
  • Verlag: Wiley
eBook (ePUB)
32,99 €
inkl. gesetzl. MwSt.
Sofort per Download lieferbar

Online verfügbar

Understanding the Predictive Analytics Lifecycle

A high-level, informal look at the different stages of the predictive analytics cycle Understanding the Predictive Analytics Lifecycle covers each phase of the development of a predictive analytics initiative. Through the use of illuminating case studies across a range of industries that include banking, megaresorts, mobile operators, healthcare, manufacturing, and retail, the book successfully illustrates each phase of the predictive analytics cycle to create a playbook for future projects. Predictive business analytics involves a wide variety of inputs that include individuals' skills, technologies, tools, and processes. To create a successful analytics program or project to gain forward-looking insight into making business decisions and actions, all of these factors must properly align. The book focuses on developing new insights and understanding business performance based on extensive use of data, statistical and quantitative analysis, explanatory and predictive modeling, and fact-based management as input for human decisions. The book includes: An overview of all relevant phases: design, prepare, explore, model, communicate, and measure Coverage of the stages of the predictive analytics cycle across different industries and countries A chapter dedicated to each of the phases of the development of a predictive initiative A comprehensive overview of the entire analytic process lifecycle
If you're an executive looking to understand the predictive analytics lifecycle, this is a must-read resource and reference guide.


    Format: ePUB
    Kopierschutz: AdobeDRM
    Seitenzahl: 240
    Erscheinungsdatum: 30.07.2014
    Sprache: Englisch
    ISBN: 9781118938928
    Verlag: Wiley
    Größe: 1295 kBytes
Weiterlesen weniger lesen

Understanding the Predictive Analytics Lifecycle

Chapter 1
Problem Identification and Definition

How executives focus resources and assess an organization's readiness for meeting the challenges posed by new business realities

Recently I met with a pair of business executives at the Gaylord Convention Center near Washington, DC.

Two analysts glided their way toward me. I smiled and went in for handshakes, exclaiming "Hello there!" Their names were Zizi and Javier. Both worked for a big corporation right outside of the Beltway in Maryland. I quickly launched into a flurry of business jargon, briskly walking toward the coffee kiosk, mouth running at a hundred million miles per minute. The executives shuffled after me, saying "We are very interested in finding out more about developing a modern analytical system."

I bought a soy latte with an extra espresso shot. As the caffeine kicked in, I started by asking, "What is your firm's level of analytical maturity?"

Javier looked at me and said, "Before we get started, do we have an NDA in place?" A nondisclosure agreement is a document signed to protect both parties. (A sample agreement is presented at the end of this chapter.) "We sure do," I answered. "Great! So let's continue."

Javier stammered, "I-I don't know. I believe that analysis is a portion of the transformation cycle from data to knowledge to wisdom. So, probably the analytical maturity of an enterprise would tell how well it can leverage analysis and close the information gap. I am not sure where I would say our company is exactly."

My eyes met his as I popped a huge sparkly smile. "Everybody knows the four key levels of an analytical framework are. . . ."

I waited for a response. Zizi replied, "Infrastructure, functionality, organization, and business, and these levels can be translated into an information evolution model for analytical applications." 1

Javier piped up, "What is the importance of this?"

I answered, "Those organizations that try simply to define and implement an advanced analytical solution in one step may end up taking far too long to finish building it and reap its benefits."

Zizi lowered her glasses and continued my thought seamlessly. "And then, most likely, the analytical solution delivered will not meet needs because requirements usually change after an initiative is initiated or because the technology has already changed. We've been through that before."

"Exactly!" I added, "There is an overarching need to build flexibility into contemporary analytical systems. Particularly now that data are growing exponentially and we are faced with big data everywhere. I believe enterprises need to assess the overall maturity of their analytical initiative and aim to add value incrementally rather than use an all-at-once approach. This is very important with the big data challenges. Results and challenges differ depending on the level of analytical maturity. I think the assessment of needs for an analytic platform or workbench should include choosing an appropriate software architecture for analysis and reporting, a hardware environment, a big data integration approach, and, of course, a data model for their structured data, among other things."

They wondered, "Is that enough to ascertain success?"

I told it to them straight. "Hey, it's anybody's guess, but it increases the probability of success significantly!"

Results usually are measured in terms of effective usage of information technology (IT) investments and improved operational efficiency. Challenges primarily occur with IT infrastructure, culture, software technology, and functionality.

They looked at each

Weiterlesen weniger lesen