Open the Simulation Page

Select a Distribution Type

• Use the dropdown labeled Select Distribution to choose the type of distribution you want to simulate.

• Options include:

  • Gaussian (2D) with equal covariance matrices
  • Gaussian (2D) with different covariance matrices
  • Normal Distribution (1D)
  • Bernoulli Distribution (1D)
  • Poisson Distribution (1D)

Learn About the Distribution

• Click the info icon (ℹ️) next to the distribution dropdown to open a popup modal that explains the selected distribution’s theory and properties.

Adjust Sample Size

• Use the Sample Size Per Class slider to control how many data points will be generated for each class (from 20 to 200).
• The current sample size value is displayed next to the slider.

Set Distribution Parameters

• Depending on the distribution you select, dynamic parameter input controls will appear in the Parameters section.
• Modify these inputs (means, covariances, probabilities, etc.) to shape the distributions according to your experiment.

Generate Data

• Click Generate Data to create random samples based on the chosen distribution and parameters.
• The generated data points will appear plotted on the canvas on the right.

Estimate Parameters (MLE)

• Click Estimate MLE to compute the Maximum Likelihood Estimates of the distribution parameters from the generated data.
• This updates the model parameters and the visualization accordingly.

Show Classifier Decision Boundaries

• Click Show Classifier to display the decision boundaries calculated from the model parameters.
• This helps visualize how the classifier distinguishes between classes.

Animate Data Points

• Click the Play button (▶️) to animate the data points being plotted one by one, helping you visually understand the data generation process.

View Performance Metrics

• The Classifier Performance panel updates dynamically with accuracy, Precision, Recall & F1 Score
• Use these to evaluate how well the classifier is performing on the generated data.