Samstag, 21. September 2019
05:20 - 05:20
Der Workshop mit Dean Abbott wird auf ENGLISCH gehalten
Die Workshopplätze sind limitiert – sichern Sie sich Ihren Platz rechtzeitig!
- Practitioners: Data Analysts who would like a introduction to data science including dozens of principles to assist in building machine learning models.
- Technical Managers: Project leaders, and managers who are responsible for leading teams of data scientists and want to understand what data scientists do.
Knowledge Level: Familiarity with the basics of statistics.
Predictive Analytics for Practitioners
Predictive analytics has moved from a niche technology used in a few industries, to one of the most important technologies any data-driven business needs. Because of the demand, there has been rapid growth in university programs in machine learning and data science. These teach the science well, but do not describe the “art” of predictive analytics, which includes practical tradeoffs when data is imperfect.
This workshop will cover the practical considerations for using predictive analytics in your organization through the six stages in the predictive modeling process as summarized in CRISP-DM:
- Business Understanding – how to define problems to solve using predictive analytics
- Data Understanding – how to describe the data
- Data Preparation – how and why to create derived variables and sample data
- Modeling – the most important supervised and unsupervised modeling techniques
- Evaluation – how to match modeling accuracy with business objectives to select the best model
- Deployment – how to use models in production
Practical tips are given throughout the workshop including:
- Which transformations of data should be used for which algorithms?
- Which algorithms match what kinds of problems?
- How does one measure model accuracy in a way that makes sense for the business?
- How does one avoid being fooled with predictive models, thinking they are behaving well when in reality they are brittle and doomed to fail?
Case studies that illustrate principles will be used throughout the workshop, drawn from Mr. Abbott’s more than 30 years of practical experience solving problems in both the private sector and public sector. The techniques are software independent, but Mr. Abbott will demonstrate solutions using an open source software package.
Every registered attendee will receive a copy of Mr. Abbott’s book “Applied Predictive Analytics”
This workshop will benefit anyone who has worked with data—whether in spreadsheets, statistics programs, R, Python, or other software—and would like to learn the practical side of predictive analytics.
All attendees will receive a course materials workbook containing more than 200 slides and an official certificate of completion at the conclusion of the workshop.
- Workshop program starts: Overview, Data Understanding
- 10:30 – 11:00
- Morning Coffee Break
- 11:00 – 12:30
- Data Preparation
- 12:30 – 13:30
- 13:30 – 15:00
- Classification Models and Model Evaluation
- 15:00 – 15:30
- Afternoon Coffee Break
- 15:30 – 17:00
- Regression Models; Model Deployment
- End of the Workshop
Dean Abbott is Co-Founder and Chief Data Scientist at SmarterHQ. Mr. Abbott is an internationally recognized expert and innovator in data science and predictive analytics, with three decades of experience solving problems in customer analytics, fraud detection and tax fraud, risk modeling, text mining, survey analysis, and many more. He is frequently included in lists of the most pioneering and influential data scientists worldwide.
Mr. Abbott is the author of Applied Predictive Analytics (Wiley, 2014, 2nd Edition forthcoming in 2020) and co-author of The IBM SPSS Modeler Cookbook (Packt Publishing, 2013). He is a popular keynote speaker and workshop instructor at conferences worldwide and serves on Advisory Boards for the UC/Irvine Predictive Analytics and UCSD Data Science Certificate programs.
He holds a B.S. in Computational Mathematics from Rensselaer Polytechnic Institute (1985) and a Master of Applied Mathematics from the University of Virginia (1987).