Implementing Analytics: A Blueprint for Design, Development, and Adoption

By Nauman Sheikh

Implementing Analytics demystifies the concept, technology and application of analytics and breaks its implementation down to repeatable and manageable steps, making it possible for widespread adoption across all functions of an organization. Implementing Analytics simplifies and helps democratize a very specialized discipline to foster business efficiency and innovation without investing in multi-million dollar technology and manpower. A technology agnostic methodology that breaks down complex tasks like model design and tuning and emphasizes business decisions rather than the technology behind analytics.

• Simplifies the understanding of analytics from a technical and functional perspective and shows a wide array of problems that can be tackled using existing technology
• Provides a detailed step by step approach to identify opportunities, extract requirements, design variables and build and test models. It further explains the business decision strategies to use analytics models and provides an overview for governance and tuning
• Helps formalize analytics projects from staffing, technology and implementation perspectives
• Emphasizes machine learning and data mining over statistics and shows how the role of a Data Scientist can be broken down and still deliver the value by building a robust development process

• Language: English
• Publisher: Elsevier Science
• Release date: May 6, 2013
• ISBN: 9780124016811

Nauman Sheikh

Nauman Sheikh is a veteran of the data architecture profession who has built dozens of large scale operational and analytical systems over the last 18 years. He has worked in three continents solving business challenges in consumer credit, risk, fraud and direct marketing areas dealing with a variety of cultural, technological and legal challenges surrounding data and its use. He is a hands-on practitioner with skills ranging from analytical reporting to data mining models to analytics driven business decisions and their audit and control frameworks.

Book preview

1

Concept

Chapter 1 Defining Analytics

Chapter 2 Information Continuum

Chapter 3 Using Analytics

Chapter 1

Defining Analytics

Contents

The Hype

The Challenge of Definition

Definition 1: Business Value Perspective

Definition 2: Technical Implementation Perspective

Analytics Techniques

Algorithm versus Analytics Model

Forecasting

Descriptive Analytics

Clustering

Predictive Analytics

Prediction versus Forecasting

Prediction Methods

Decision Optimization

Conclusion of Definition

The Hype

Analytics is one of the hot topics on today’s technology landscape (also referred as Big Data), although it is somewhat overshadowed by the high-profile social media revolution and perhaps also by the mobile revolution led by Apple Inc., which now includes smartphones, applications, and tablets. Social media, mobile, and tablet revolutions have impacted an individual’s life like never before, but analytics is changing the lives of organizations like never before. The explosion of newer data types generated from all sorts of channels and devices makes a strong argument for organizations to make use of that data for valuable insights. With this demand and emergence of cost-effective computing infrastructure to handle massive amounts of data, the environment is ripe for analytics to take off. However, like any technology that becomes a buzz word, the definition becomes more and more confusing with various vendors, consultants, and trade publications taking a shot at defining the new technology (analytics is actually not new but it has been reborn with Big Data; see Chapter 11). It becomes extremely difficult for people intrigued by this topic to sort through the confusing terminology to understand what it is, how it works, and how they can make use of it. This happened with ERP and e-commerce in the mid- to late 1990s and with CRM in the early 2000s. Over time, as the industry matures, consensus emerges on what is the definition and who are the dominant players, respectable trade publications, established thought leaders, and leading