Abstract

The project goal is to showcase the benefits and powerfulness of Bayesian statistics. Here, I introduce you four scenarios, weaving one or more Bayesian subtopics, such as Bayesian Inference , Bayesian Classification, Bayesian Survival Analysis, et al.

These scenarios are structures as narrative stories to guide the portfolio’s layout:

  1. starting with a real-world problem introduction,
  2. moving through analysis steps,
  3. incorporating interactive visualizations
  4. ending with insights and reproducibility notes.

Each scenario uses Python for code analysis (e.g., PyMC or Stan via pymc or cmdstanpy), public datasets and GitHub for hosting code/notebooks alongside a dynamic webpage(i.e, via Streamlit or Dash embedded in HTML for interactivity)

The detailed roadmap of each scenarios are presented in the table below. For each, I’ve outlined 5-7 key steps with descriptions, including where interactive visualizations fit in. This layout ensures a logical flow: problem → method → visualization → results → reflection.

Data