Statistical Methods in MATLAB


MATLAB Fundamentals and knowledge of basic statistics.
Day 1 of 2
Importing and Organizing Data

Objective: Understand the import methods and data types available in MATLAB and Statistics and Machine Learning Toolbox to bring data into MATLAB and organize it for analysis. Perform common tasks, such as merging data and dealing with missing data.

  • Importing data
  • Data types
  • Dataset arrays
  • Merging data
  • Categorical data
  • Missing data
Exploring Data

Objective: Perform basic statistical investigation of a data set, including visualization and calculation of summary statistics.

  • Plotting
  • Central tendency
  • Spread
  • Shape
  • Correlations
  • Grouped data

Objective: Use the functionality in Statistics and Machine Learning Toolbox to investigate different probability distributions and to fit distributions to a data set.

  • Probability distributions
  • Distribution parameters
  • Comparing and fitting distributions
  • Nonparametric fitting
  • Distribution objects
Hypothesis Tests

Objective: Use Statistics and Machine Learning Toolbox to determine how likely an assertion about a data set is. Apply hypothesis tests for common uses, such as comparing two distributions and determining confidence intervals for a sample mean.

  • Hypothesis tests
  • Tests for normal distributions
  • Tests for nonnormal distributions
Day 2 of 2
Analysis of Variance

Objective: Use Statistics and Machine Learning Toolbox functions to compare the sample means of multiple groups and to find statistically significant differences between groups.

  • Multiple comparisons
  • One-way ANOVA
  • N-way ANOVA
  • Nonnormal ANOVA
  • Categorical correlations

Objective: Perform predictive modeling by fitting linear and nonlinear models to a data set. Explore techniques for improving model quality.

  • Linear regression models
  • Fitting linear models to data
  • Evaluating the fit
  • Adjusting the model
  • Logistic and generalized linear regression
  • Nonlinear regression
Working with Multiple Dimensions

Objective: Become familiar with techniques for reducing the dimensionality of a data set, and perform classification of categorical responses.

  • Feature transformation
  • Feature selection
  • Classification
  • Clustering
Random Numbers and Simulation

Objective: Use random numbers to evaluate the uncertainty or sensitivity of a model, or perform simulations. Generate random numbers from various distributions, and manage the MATLAB random number generation algorithms.

  • Bootstrapping and simulation
  • Generating numbers from standard distributions
  • Generating numbers from arbitrary distributions
  • Controlling the random number stream