The Business Intelligence Blog

Slicing Business Dicing Intelligence

101 Ways to Sabotage Your Predictive Analytics Project  

And there is the first five here. Interesting Read, I liked it.

The strategic approach and project design for predictive analytics is substantially different than the other areas of business intelligence. Unlike a data warehouse design, which is similar to an engineering project, predictive analytics and data mining are a discovery process. And while several consortiums have standardized formal processes to accommodate discovery and iterative process, the practice remains riddled with common pitfalls.

Those who make the effort to educate themselves on the industry-standard approach to predictive analytics are nearly assured to reap residual returns – long before their counterparts who typically rush to acquire a tool and dive headlong into the data.

share the post:
  • Twitter
  • Facebook
  • LinkedIn
  • FriendFeed
  • del.icio.us

The article has

6 responses

Written by Guru Kirthigavasan

May 18th, 2010 at 6:22 pm

6 Responses to '101 Ways to Sabotage Your Predictive Analytics Project'

Subscribe to comments with RSS or TrackBack to '101 Ways to Sabotage Your Predictive Analytics Project'.

  1. I liked it. good read

    Kuldeep

    8 Jun 10 at 5:28 am

  2. As I see it, most “ordinary” business intelligence projects would also benefit from a more experimental development approach.

    However, the reason is slightly different.

    In many organizations there is a lack of experience and skills withing business intelligence. Thus, making some small prototypes fast will increase the learning curve drastically and ensure a better design and architecture in the end.

  3. Our ExisOne BI is available for use on either your owned computer(s) or on hosted service. It is a full featured BI solution for enterprises as well as medium to small businesses. All editions have identical features. Please take a look at our website http://www.exisone.com. We have a live demo for you to try it out. ExisOne can be implemented in as little as one week. And, it is so intuitive that a user can learn to use it in minutes.

    exisone

    7 Jul 10 at 7:37 pm

  4. very nice & knowledgeable article…must read to two time…

    hukrockport

    8 Jul 10 at 3:37 am

  5. Hello,

    SiliconIndia is organizing Business Intelligence Conference on July 31, 2010 (Saturday) in Bangalore.

    At the SiliconIndia Business Intelligence Conference, you will learn about the challenges, solutions and future of BI from top industry experts and leading-edge peers. There will be exciting technical sessions, delivered by the people who know BI best – the experts who design, build and support the products, the partners and consultants who specialize in deploying Business Intelligence solutions, and the customers who use BI products each and every day. You’ll experience a complete technical immersion, shared with a developer community that is passionate about all exciting developments in the business intelligence world.

    Technical TRACK
    Emerging Trends and Technologies in the Business Intelligence Market
    BI Design and Implementation Challenges
    Integrating Search and Business Intelligence
    Design approaches for real-time BI
    Best Practices for Data Warehousing
    Building Interactive Dashboards
    Building a BI Solution Leveraging Analytical Reporting
    Creating and Defining Customized Analytical Functions
    Analyzing Data with Advanced Visualizations
    Business Intelligence for the Mobile Workforce
    Open-source business intelligence (BI) tools

    Customer Business Value TRACK
    This track will highlight case studies from enterprises using BI products, sharing lessons learned in deploying business intelligence applications in their organizations. Learn how leading companies use BI software to improve operations, reduce inefficiencies, and enhance business performance. Case Studies: Healthcare, Banking, Telecom, Manufacturing, Insurance, Retail, Financial Services, Utilities (Power), Technology Sector, Media sector

    Top Reasons to Attend
    . Hear from leading minds in the industry
    . Take a closer look at new tools, platforms and technologies
    . Objective information on BI strategies, techniques, and technologies
    . Learn development techniques & tricks
    . Meet the experts and network with the BI community
    . Fresh ideas and inspiration you can use immediately back at the office

    DATE: July 31st, 2010 (Saturday)
    TIME: 8.30 AM to 5 PM
    VENUE: Nimhans Convention Center, Bangalore

    TO REGISTER NOW:
    You can send your nominations by responding to this mail or Call Akanksha , Events Executive on 080 43402053
    We have limited Seats. Attendance is limited to maintain an intimate setting and foster dialogue among all participants. There will be a registration fee of Rs 500 only. This is towards: Access to Sessions, Tea/ Coffee & Refreshments and Lunch.

    We will also encourage you to pass on this note to your colleagues so that they could benefit from this conference. Please note: You may also ask your organization/ manager to sponsor you and your colleagues to attend this exciting conference.

    You don’t want to miss a power packed day with several BI developers, architects, project managers and teams, IT managers and professionals, analysts and business decision makers coming together to build their BI expertise.

    Regards,
    Akanksha
    Events Executive
    silicon india
    080 43402053

    Akanksha

    12 Jul 10 at 2:23 am

  6. The Secret to Predictive Analytics is the star schema turned inside out. Leave the data in the transactional systems. Turn the fact table into an RDF triple store that contains an integrated set of RDF patterns linking the KEY predictive TAGS from each of the transactional systems. You need a harmonized vocabulary, thus the role of Master Data Management. You should only need about 14 – 21 concepts and six levels of weights and ranks to get the computational facets you need to discover historical patterns, map those to current state patterns, and extend the results to future probabilities using SPARQL patterns with known constraints. Predicting the future is quite easy when you understand the semantics of human cognitive models.

    Art Conroy

    29 Aug 10 at 6:24 am

Leave a Reply