Central Time CT
Co-Founder/Chief Data Scientist
Smarter HQ, Inc.
At its core, data science leverages machine learning and statistics-based algorithms to find patterns in data. The second day of the TDWI Data Science Bootcamp covers the most commonly used algorithms in data science, providing overviews of what they are and how they find patterns—without an in-depth treatment of the mathematics.
You will learn how to match these common algorithms to analytics objectives and best practices to ensure they lead to business results. We will explore similarities and differences between the algorithms, what types of patterns each can find most easily, and which patterns are more difficult for each to uncover. You will learn how data preparation and feature creation (from day one) influence the accuracy of your results.
Explanations of model accuracy will cover standard metrics, such as mean-squared error and percent correct classification, as well as other metrics that are more useful in practice. Analytic models can only lead to business impact if they are trusted by stakeholders who are willing to act on their results. You will learn how to explain models and model accuracy to business stakeholders. Model interpretation strategies and metrics for complex algorithms will be also be described, equipping you with the communication techniques needed to generate business value.
You Will Learn
- Supervised learning techniques including decision trees, regression models, and neural networks
- What types of pattern each technique is good at identifying, and how to match data science techniques to your analytics objectives
- The fundamentals of basic algorithms including k-nearest neighbor, naïve Bayes, and support vector machines
- Unsupervised learning techniques for clustering
- Evaluation techniques for supervised and unsupervised learning models
- Analytics practitioners
- Data scientists
- Business analysts
- Data engineers and data pipeline developers
- Project leaders, technical managers, and engineers