Home / Collections / Courses for Executives: Course Catalog / Predictive Analytics for Business Analysts: Using AI to Drive Strategy

Predictive Analytics for Business Analysts: Using AI to Drive Strategy

$9,280.00
Predictive Analytics for Business Analysts: Using AI to Drive Strategy

Map - NYU Stern School of Business

Location

NYU Stern School of Business
44 West 4th Street
New York, NY 10012

View Map

Predictive Analytics for Business Analysts: Using AI to Drive Strategy

$9,280.00

has been successfully added to your cart . Feel free to continue shopping or check out .

Overview

NYU's Center for Data Science and the NYU Stern School of Business bring you this course on predictive analytics, which is the key source of business value from artificial intelligence. Almost all products and services can be augmented with predictions, and as predictions become more accurate and less expensive, business leaders are redesigning their companies to take advantage of these new capabilities. Until recently, analysts needed software engineering skills to develop predictive models and to put them into operation. But a new wave of technology -- Automated Machine Learning -- has made the power of predictive analytics accessible to many decision-makers and executives.

In this course, participants will use the Microsoft Azure automated machine learning tool to build predictive models that help inform valuable business decisions. We will discuss the fundamentals of predictive analytics, learn about what kind of predictions can be made, and uncover the data is needed to make these decisions. After building various predictive models, you will then make the leap from analysis to strategy and learn how your predictions can help redesign existing product and services offerings to meet customer needs, and conceptualize new offers and business opportunities. This course combines lectures, hands-on labs and case studies to enhance learning. Because it teaches the principles underlying creation of predictions, you will be able to quickly learn to apply those principles to tools similar to the Azure tool.

Joint Expertise from NYU Center for Data Science and NYU Stern

This course is a joint offering between the NYU Center for Data Science and NYU Stern School of Business, and draws on the strengths of each. The Center for Data Science serves as the focal point for NYU's university-wide efforts in Data Science and is helping to shape this new field. For more than 100 years, NYU Stern has been inspiring executives to innovate, lead and transform themselves and their organizations.

Certificates and Credits

Upon completion of this course, participants will receive a Certificate of Achievement.

Program Takeaways

  • Fundamentals of Predictive Analytics

    Developing a sound understanding of the principles of predictive analytics in order to engage with data scientists and business experts in your organization to conceptualize, build, and deploy predictive models.
  • Prediction Models

    Analyze existing data sets and generate several models that can make useful predictions, including customer churn predictions, the probability that order fulfillment processes will fail, and which customers are likely to accept offers.
  • Testing Model Accuracy

    Determine the accuracy of models by testing them with new data once the models are developed, and practice identifying the model(s) that should be operationalized within the company based on this assessment.
  • Deploying Models

    Recognize the need to build decision-centric organizations to get full value from their investments in analytics

Who Should Attend

Although there are no formal education or background requirements, this course is designed for executives who meet the criteria below. While we strongly encourage global participation, please note that all courses are taught in English. Proficiency in written and spoken English is required.

  • Years of Experience

    Designed for professionals with 5+ years of work experience
  • Job Functions

    Ideal for professionals with a variety of titles such as business analyst, marketing manager, product manager, solution designer, and business architect.
  • Prerequisites

    Intended for individuals who are familiar with using spreadsheets and presentations to understand the opportunities a business faces and to develop consensus around how to change it.

Agenda

The following agenda is a sample and subject to change.

8:30 am - 9:00 am: Breakfast

Lecture Session: Understanding the Basics of Predictive Analytics and Machine Learning

  • Terminology and basics: data science, big data, AI, machine learning, supervised learning, unsupervised learning, regression vs. classification, predictions vs. decisions
  • The data science process: the CRISP-DM process, how it differs from software development processes, Agile project management methods
  • How machines can learn: fitting a custom-designed linear regression model by hand, model design choices, framework for specifying a model

12:15 pm - 1:15 pm: Lunch

Lab Session: Getting Started with Predictive Analytics

  • Establishing access to the cloud service, navigating the screens, reviewing and summarizing data
  • Designing and training a linear model, interpreting the output

Project Session: Introduction to the Problem

  • Overview of the course project
  • Explanation and exploration of the project dataset

4:30 pm - 5:00 pm: Day 1 Conclusion and Evaluations

8:30 am - 9:00 am: Breakfast

Lecture Session: Evaluating Regression Models

  • Decision trees for regression: splitting nodes, interpreting the results, decisions trees and equivalent rule sets.
  • Overfitting: what it is, how it can happen, how to avoid it, validation vs. training data, cross validation when training data are limited, regularization, the “true” model vs. the model we use; hyperparameters and learned parameters; estimated the error on data we have not seen.
  • Getting value from predictions: the business setting, the prediction, the decision enabled in part by the prediction, financial impact of the improved decision.

12:15 pm - 1:15 pm: Lunch

Lab Session: Using Additional Regression Models

  • Decision trees for regression, other regression models
  • Model assessment, validation and cross validation

Project Session: Independent Group Work

  • Participants work in groups of 3 to 5 to use predictive analytics to find a solution to the project problem

4:30 pm - 5:00 pm: Day 2 Conclusion and Evaluations

8:30 am - 9:00 am: Breakfast

Lecture Session: Predicting Classes

  • Classification vs. regression: predicting probabilities, why regression will not work for classification, decision trees for classification, measuring the amount of information using entropy.
  • Logistic regression for classification: model form, interpreting the coefficients, confusion matrix, threshold selection, ROC
  • Case study: implementation and deployment of a predictive model: what happened after the model was put into operations.

12:15 pm - 1:15 pm: Lunch

Lab Session: Implementing Classification Models

  • Classification methods: logistic regression, decision trees, other classification models
  • Evaluation of classifiers: ROC, confusion matrix

Project Session: Independent Group Work

  • Participants work in groups of 3 to 5 to use predictive analytics to find a solution to the project problem

4:30 pm - 5:00 pm: Day 3 Conclusion and Evaluations

8:30 am - 9:00 am: Breakfast

Lecture Session: Understanding Advanced Models and Redesigning Businesses to Leverage Predictions

  • Advanced predictive models: neural networks, deep learning, random forests.
  • Learning to predict associations: clustering, K means, text as data
  • Business strategy and cheap, accurate predictions: analyses of value added sequences, digitization of processes, recomposing offerings to leverage predictions, building data science capabilities, case study

12:15 pm - 1:15 pm: Lunch

Lab Session: Implementing Advanced Models

  • Additional models: neural networks, processing images, random forests
  • Additional capabilities of the Auto ML tool: time series forecasting, ensemble models, rebalancing training data, deployment into operations

Project Session: Group Project Workshop with Instructor

  • Groups continue to work on project and meet with instructor for guidance and troubleshooting

4:30 pm - 5:00 pm: Day 4 Conclusion and Evaluations

8:30 am - 9:00 am: Breakfast

Project Session: Independent Group Work

  • Participants work in groups of 3 to 5 to finalize their projects
  • Finalize the modeling
  • Create a short presentation to share with your classmates and instructor

12:15 pm - 1:15 pm: Lunch

Project Session: Group Project Presentations & Case Solution

  • Groups present their projects and explain the solutions they developed
  • Instructor provides feedback on group projects and reveals how the company in the project actually used predictive analytics to solve their problem

4:30 pm - 5:00 pm: Program Conclusion and Evaluations