Predictive Analytics audiobook cover - The Power to Predict Who Will Click, Buy, Lie, Or Die

Predictive Analytics

The Power to Predict Who Will Click, Buy, Lie, Or Die

Eric Siegel

★ 4.0 / 5(87 ratings)
Start ListeningDownload App
Categories:
Marketing & Sales

If You're Curious About These Questions...

You should listen to this audiobook

Listen to Predictive Analytics — Free Audiobook

Loading player...

Key Takeaways from Predictive Analytics

Learning Tools

Reinforce what you learned from Predictive Analytics

Mind Map

Predictive Analytics
Core Concepts & Functionality+
Ethics, Privacy & Prejudice+
Data Quality & Balance+
Machine Learning Risks & Rewards+
Ensemble Models & Crowdsourcing+
Natural Language Processing+
The Uplift Model & Persuasion+

Quiz — Test Your Understanding

Question 1 of 8
What is the primary purpose of predictive analytics (PA) in a business context, according to the text?
  • A. To determine which advertisement has the broadest general appeal.
  • B. To guarantee the success of expensive marketing campaigns.
  • C. To predict the likeliest responses of specific individuals to specific situations.
  • D. To replace human decision-making with automated algorithms.
Question 2 of 8
How does 'backtesting' improve the accuracy of predictive tools?
  • A. By running the model backward to find programming errors.
  • B. By feeding old data into the model to see if it accurately predicts past outcomes.
  • C. By surveying customers after a marketing campaign has finished.
  • D. By comparing a company's data with its competitors' historical data.
Question 3 of 8
Why does the text suggest that using predictive analytics in the criminal justice system can be problematic?
  • A. It requires more data than most cities are able to legally collect.
  • B. It can inadvertently lead to racial profiling by heavily weighing factors like a convict's zip code.
  • C. It violates a convict's right to privacy by accessing their social media accounts.
  • D. It is generally less accurate than traditional policing methods.
Question 4 of 8
According to the text, what is a primary cause of false correlations in predictive analytics, such as the claim that orange cars are less likely to be faulty?
  • A. Using old, outdated data.
  • B. Relying entirely on social media and blog posts.
  • C. Using unbalanced data sets with too many variables in one area.
  • D. Human bias entering the programming phase.
Question 5 of 8
What is the danger of 'overlearning' in machine learning models?
  • A. The computer requires too much processing power and crashes.
  • B. The model begins to delete older, valuable data to make room for new data.
  • C. The model finds meaningless correlations and makes faulty predictions based on them.
  • D. The model becomes entirely autonomous and ignores human input.
Question 6 of 8
How did the 2008 Netflix crowdsourcing competition significantly advance the field of predictive analytics?
  • A. It proved that single predictive models are more reliable than combined ones.
  • B. It demonstrated the power of the 'ensemble model,' which combines multiple predictive models.
  • C. It showed that human intuition still outperforms machine learning in predicting movie tastes.
  • D. It highlighted the critical shortage of analytical skills in the tech industry.
Question 7 of 8
How did IBM's Watson differ from traditional predictive analytics models?
  • A. It was built to predict future outcomes based on financial data.
  • B. It relied entirely on a single, highly advanced predictive model.
  • C. It was designed to eliminate possibilities and predict the likeliest answer rather than predict a future event.
  • D. It was programmed to understand human sarcasm with 100 percent accuracy.
Question 8 of 8
What specific problem does the 'uplift model' attempt to solve in marketing?
  • A. Determining how to reach the maximum number of people with a single advertisement.
  • B. Identifying which customers will be positively persuaded by an ad, rather than ignoring or being annoyed by it.
  • C. Calculating the exact financial return on investment for a social media campaign.
  • D. Predicting which competitors are most likely to steal a company's target audience.

Predictive Analytics — Full Chapter Overview

Predictive Analytics Summary & Overview

Predictive Analytics (2016) provides a helpful introduction to a complex and fascinating field. Learn how data gets crunched so that people can make more informed decisions, a practice that has drastically altered the way the world conducts its research and runs its businesses. Siegel offers an enlightening glimpse at the wide-ranging areas that have been forever changed, from marketing to health care, banking to artificial intelligence.

Who Should Listen to Predictive Analytics?

  • Business students interested in applied analytics
  • Readers interested in economics
  • Tech geeks curious about artificial intelligence

About the Author: Eric Siegel

Eric Siegel is a world-renowned leader in the field of predictive analytics and the founder of the Predictive Analytics World Conference Series. A former Columbia University professor, he’s also the executive editor of the Predictive Analytics Times.

© Eric Siegel: Predictive Analytics copyright 2016, John Wiley & Sons Inc. Used by permission of John Wiley & Sons Inc. and shall not be made available to any unauthorized third parties.

🎧
Listen in the AppOffline playback & background play
Get App