Applied Machine Learning

collapse

Course Details

About this Course


Explore the iterative process necessary to solve real-world business problems through the application of machine learning. Learn to begin with a concise statement of the underlying business needs, identifying and sourcing raw data for analysis and modeling. Continue with data preparation, including data cleaning and creation and selection of features, followed by model learning, evaluation, scoring and implementation. This is an iterative process in which insights gained during model development and evaluation may motivate going back to the beginning to refine the problem and/or source additional data, add or delete features, etc. Explore this iterative process by leveraging tools of the trade, including open source tools.

Topics include:

  • Common terms and techniques (e.g., model vs. learner, predictive model, operational model, data set, event/instance, supervised and unsupervised learning, train (fit), validate (select), cross-validation)
  • Business needs assessment, data identification and data sourcing
  • Data cleansing, feature engineering, feature exploration and feature selection
  • Machine learning (model learning), model evaluation and validation
  • Model testing (historical testing, live A/B testing)
  • Model scoring and implementation, measurement and feedback
  • Tools of the trade (open source, MS environment)
  • Put a model into production, discuss what you need to monitor and determine when to retrain

Program Overview

Complete the courses listed below to earn the certificate. You may be able to take individual courses without enrolling in the certificate program; check the course pages for details.