Software Engineering Institute

Through the fundamentals of statistics related to cybersecurity, aspiring data scientists can:

• Gain knowledge of common problems that a data scientist encounters
• Become fluent in statistics with the help of a scripting language
• Increase predictive power and reduce risk within a model
• Better estimate parameters for a dataset
• Investigate and solve problems in the cybersecurity realm

Please note that successful completion of this course is a required component of the CERT Applied Data Science for Cybersecurity Professional Certificate. To learn more about the Professional Certificate and discounted package pricing, please go to: SEI Certificates.

• Those with a particular interest in data science and cybersecurity, but limited experience with both concepts.

After successful completion of this course, you will:

• be able to perform appropriate hypothesis tests for corresponding data and recognize different distributions of data
• be able to recognize relevant learning methods for a certain dataset
• have an appreciation for probability being at the forefront of statistics, model selection, and data science as a whole
• explain why bias and variance are so critical to assessing model performance
• explain the differences between types of learning methods and regularization techniques
• explain the concept of risk within estimation
• complete tasks involving model selection
• complete tasks involving finding optimal parameters for a model, distribution, or set of data

In this course, students will learn about and investigate statistical techniques relied upon in the cybersecurity realm. These include:

• fundamentals of probability, including:
  • basic properties of probability
  • common distributions of data
  • visualizing data
  • maximum likelihood estimation
  • hypothesis testing
  • parametric and nonparametric tests
  • supervised and unsupervised learning methods
  • the bias-variance tradeoff
  • training and validating a model
  • regularization techniques for creating more generalizable models.

These concepts will be exercised in labs involving density and maximum likelihood estimation, hypothesis testing with z-tests, linear regression, and logistic regression.

This course is presented in the form of video instruction presented by experts from the SEI CERT Division. Downloadable materials include course presentation slides, instructions for lab exercises, jupyter license, and instructions for using a jupyter notebook. Learners will also be able to access additional resources related to the subject matter.

Learners should have fundamental knowledge of descriptive statistics and a working knowledge of a programming language (preferably Python or R). A working knowledge of calculus and linear algebra is helpful.

To access the SEI Learning Portal, your computer must have the following:

• For optimum viewing, we recommend using the following browsers: Microsoft Edge, Mozilla Firefox, Google Chrome, Safari
• These browsers are supported on the following operating systems: Microsoft Windows 8 (or higher), OSX (Last two major releases), Most Linux Distributions
• Mobile Operating Systems: iOS 9, Android 6.0
• Microsoft Edge, Firefox, Chrome and Safari follow a continuous release policy that makes difficult to fix a minimum version. For this reason, following the market recommendation we will support the last 2 major version of each of these browsers. Please note that as of January 2018, we do not support Safari on Windows.