Just because we have amassed a huge amount of data doesn’t mean that we really understand what it is telling us. To move from raw data to actionable information, we frequently must use algorithmic techniques. However, the ever-growing range of available algorithms and the confusing landscape of algorithmic technologies can make it hard to select, apply, and integrate algorithmic intelligence into your data analysis pipeline.
This is a two-part course that can be taken together as a full-day course, or each part can be taken individually without the other. In this first part we will develop a comprehensive framework with which to categorize and assess algorithms, look at a number of real-world case studies in which algorithms dramatically changed the ways in which visualization was leveraged, and demystify some intimidating advanced algorithmic techniques like decision trees and deep learning neural networks. We’ll touch briefly on the implementation details that are involved in developing systems that leverage algorithms but will leave most of the technical details of implementation to Part 2.
You Will Learn
- The difference between data, information, and knowledge, and how to improve visualizations with algorithms that move from data to knowledge.
- How algorithms can make the difference between having a data-driven message fall flat or driving stakeholders to action.
- A framework for thinking about algorithms and how general-purpose algorithms for numerical data can be applied to non-numeric and specialized data
- How to move from simple single-variable algorithms to two-variable algorithms to n-dimensional algorithms capable of taking many different variables into account.
- How algorithms can be combined to add value by looking at a case study in the airline industry.
- How principal component analysis, clustering, decision trees and neural networks work, and when to apply them.
- The difference between unsupervised and supervised machine learning, and how the different approaches work together.
- Some of the caveats when applying algorithms and implementing predictive analytics.
- Basic implementation factors to be aware of when developing algorithm-based solution architectures.
- Data analysts, business analysts, business intelligence professionals, analytics professionals, data scientists, and data visualization practitioners; developers or architects responsible for integrating disparate technologies; anyone responsible for finding and communicating knowledge derived from data