Master Python, statistics, machine learning, model evaluation, storytelling, and end-to-end data science portfolio projects.
Complete the Python, Math, and Data Analysis branches before moving ahead.
Use Python, notebooks, Pandas, NumPy, Matplotlib, Seaborn, environments, and clean project structure.
Learn distributions, sampling, hypothesis testing, confidence intervals, correlation, regression, and experiment design.
Profile datasets, handle missing values, find outliers, visualize trends, and write useful analysis notes.
Analyze a public dataset, clean it, visualize key patterns, and publish a short notebook with business insights.
Complete the Machine Learning Core branches before moving ahead.
Train regression and classification models, split data correctly, tune hyperparameters, and compare baselines.
Use clustering, dimensionality reduction, anomaly detection, embeddings, and similarity search for discovery tasks.
Compare three models on the same dataset, inspect errors, check leakage, and explain metric tradeoffs.
Build a baseline and improved model, document preprocessing, evaluate fairly, and write recommendations.
Complete the Feature Engineering, Deep Learning, and AI branches before moving ahead.
Create numeric, categorical, text, time, and aggregate features while avoiding leakage and unstable signals.
Understand neural networks, tensors, training loops, overfitting, embeddings, transfer learning, and GPU workflows.
Check fairness, drift, interpretability, privacy, documentation, and when not to use a model.
Train a model, explain top drivers, test subgroup performance, and write a non-technical model risk summary.
Complete the Portfolio and Production Readiness branches before moving ahead.
Track experiments, version data, package models, serve predictions, monitor drift, and document model lifecycle decisions.
Turn analysis into clear recommendations, visuals, assumptions, limitations, and next actions.
Publish a complete project with dataset, EDA, model, evaluation, explainability, deployment demo, and README.
Create a simple model API or Streamlit app with tracked experiments, clean README, and a short project write-up.