Corporate Training By Zack Academy - Building the Analytics-Driven Operation - Development

Analysts whose responsibility is focused solely on applying statistical and machine learning methods to arrive at valid prescriptive insights that align with organizational priorities may attend this vendor-neutral exposure to the Model Phase as a stand-alone course.

Building the Analytics-Driven Operation | Development course is the only class of its kind that places analytic model development directly in the context of a broader enterprise-level implementation framework. This course is not just about the mechanics of building a predictive model – it’s about building models that drive tangible, measurable results for the direct benefit of the organization. This event covers the Fourth Phase of the Seven-Phase Modeling Practice Framework: Model.

Attendees will directly experience the natural messiness and realities of organizational analytic implementation through work-along labs and guided discussions that build upon a highly realistic overall project. No other course conveys this level of knowledge transfer, experiential engagement and skill reinforcement.

Who Should Attend Building the Analytics-Driven Operation | Development Course?

• Data Scientists
• Functional Analytic Practitioners
• Big Data Analysts
• Project Leaders
• Business Analysts
• IT Professionals
• Anyone Overwhelmed with Data and Starved for Actionable Insights

Upon successful completion, students will earn 18 INFORMS Professional Development Units.

Prerequisites

While this course is designed to be taken independently, it is helpful to understand its place and function within the overall ‘The Predictive Analytics Operation’ comprehensive course.

Prior education or experience in analytics or statistics is helpful, but not required. Those seeking a deep drill-down into the mathematical or theoretical underpinnings of machine learning algorithms should refer to available academic or on-demand online offerings. The analytic algorithms in this course are actively demonstrated and conveyed from a functional perspective.

Registrants will be required to view a three-hour online “Core Concepts” orientation prior to attending this event. Access details for the Core Concepts orientation will be shared with participants prior to the start of the course.

Participants need only supply a laptop computer with Microsoft Excel. Instructions on how to download lab data and any analytic tools will be provided in a preparatory email. The instructor can assist attendees with any preparation during breaks, and before or after class.

What Makes This Seminar Unique

This course does not restrict or skew the presentation of machine learning methods through a single product. Rather, the Model Development course gives broad consideration of predictive analytics from a purely vendor-neutral perspective.

Live modeling demonstrations will precede follow-along exercises. Participants will directly experience the natural messiness of data mining to discover what really works, as well as what doesn’t and why. The instructor will show how to evaluate various features and available products based upon strengths, limitations, function, value and general performance.

As active consultants on large organization analytic implementations, each member of TMA’s seasoned faculty possesses a wealth of practical experience in applying predictive analytics across industries. This course, like no other, insists upon making predictive analytics purposeful, measurable and actionable in a larger and more complex organizational setting.

Syllabus

Core Concepts

• Prerequisite Three-Hour Preparatory Orientation
• View the full Core Concepts Topic Outline

Model Development Introduction

• Current Trends in Analytic Modeling, Data Mining and Machine Learning
• Algorithms in the News: Deep Learning
• The Modeling Software Landscape
• The Rise of R and Python: The Impact on Modeling and Deployment
• Do I Need to Know About Statistics to Build Predictive Models?
• Strategic and Tactical Considerations in Choosing a Modeling Algorithm
• What is an Algorithm?
• Is a “Black Box” Algorithm an Option for Me?

The Tasks of the Model Phase

• Generate Test Design
• Train-Test Validation
• Accept or Reject Modeling Parameters
• Test / Test / Validate
• Optimizing Data for Different Algorithms
• Build Models
• Classification
• Issues Unique to Classification Problems
• Why Classification Projects are So Common
• An Overview of Classification Algorithms
• Logistic Regression
• Neural Networks
• Naïve Bayes Classification
• Support Vector Machines
• Decision Trees
• Ensemble Methods
• Value Estimation and Regression
• Clustering
• Association Rules
• Other Modeling Techniques
• Times Series
• Text Mining
• Factor Analysis
• Model Assessment
• Evaluate Model Results
• Check Plausibility
• Check Reliability
• Model Accuracy and Stability
• Lift and Gains Charts
• Modeling Demonstration
• Assess Model Viability
• Select Final Models
• Why Accuracy and Stability are Not Enough
• What to Look for in Model Performance
• Exercise Breakout Session
• Create & Document Modeling Plan
• Determine Readiness for Deployment
• What are Potential Deployment Challenges for Each Candidate Model?
• Exercise Breakout Session and Guided Project Discussion

Wrap-up and Next Steps

• Supplementary Materials and Resources
• Conferences and Communities
• Get Started on a Project!
• Options for Implementation Oversight and Collaborative Development