IBM IBSC 2026-04-08 | Week 8 Recap + Week 9 Preview | Advanced ML, MLOps & Capstone Kickoff

Lecture recap and preview for the IBM Data Science Professional Certificate / Bootcamp - IBSC 2026-04-08 cohort. In this video, we review Week 8, where we continued the machine learning unit and moved beyond linear models into more advanced algorithms, evaluation workflows, and model deployment. We then preview Week 9, which officially begins the IBM Data Science capstone project. Week 8 focused on expanding your machine learning toolkit and thinking more realistically about how models are built, evaluated, selected, and eventually deployed. In this recap, we cover: • Decision trees and tree-based models • Ensemble methods including Random Forests, XGBoost, CatBoost, and AdaBoost • Bias, variance, model complexity, and explainability tradeoffs • Multi-class classification strategies • Support Vector Machines and K-Nearest Neighbors • Unsupervised learning with clustering • Dimensionality reduction with PCA and t-SNE • Train/test splits, validation, and cross-validation • Avoiding data leakage and overfitting • Model evaluation beyond accuracy • Rainfall prediction classification example • Comparing Logistic Regression, Random Forest, XGBoost, SVM, and KNN • Model persistence with Pickle • Production concepts including APIs, Docker, MLOps, monitoring, and drift We also preview Week 9, which begins the capstone project. In the capstone, learners act as data scientists working for Space Y and analyze SpaceX launch data to predict whether a booster can successfully land and be reused. In Week 9, the focus shifts to: • Data collection • API requests • Web scraping • Data wrangling • Exploratory data analysis • Saving and organizing notebook outputs • Preparing capstone work for GitHub • Starting the final project workflow Important note: the SpaceX API used in one of the Week 9 notebooks is currently unavailable, so students should continue with the remaining notebooks and use the provided static outputs where needed. This issue will not prevent completion of the capstone. The big idea this week: machine learning is not just about training a model. A strong data scientist also needs to evaluate models honestly, avoid workflow mistakes, understand deployment basics, and communicate results clearly in a final project. GitHub materials: https://github.com/ABoothInTheWild/ib... #ibm #datascience #machinelearning #mlops #modeldeployment #python #scikitlearn #xgboost #randomforest #bootcamp #capstone #dataanalytics #aiengineering #llmops #docker #apis