← Back to projects
Pythonscikit-learnTensorFlowDockerKubernetes

Machine Learning Zoomcamp Coursework

My completed coursework from the Machine Learning Zoomcamp: regression, classification, and neural networks, taken all the way through to deployed models on Kubernetes.

LearnerSelf-directed · 20243 min read
9 modules completed
Regression → deep learning
Deployed models with Flask & Kubernetes
Notebooks organized by module

Overview

The Machine Learning Zoomcamp is a free, project-based course from DataTalksClub that walks through the full lifecycle of a machine learning project, from framing a problem and training a first model to shipping it as a running service. This repository is my completed coursework: the notebooks, scripts, and deployment work I built while working through the camp end to end.

It isn't a single polished application so much as a record of the progression: each module builds on the last, and the repo is organized into module-1 through module-6 directories that group the work as the course moved from fundamentals into deployment and serving.

What's Covered

ModuleTopic
1Introduction to Machine Learning
2Machine Learning for Regression
3Machine Learning for Classification
4Evaluation Metrics
5Deploying ML Models
6Decision Trees & Ensemble Learning
7Neural Networks & Deep Learning
8Serverless Deep Learning
9Kubernetes & TensorFlow Serving

The arc is deliberate: start with the framing and a linear regression model, move into classification and the metrics that tell you whether a model is actually good, then take a trained model and put it behind a real interface. The later modules push deployment further: containers, serverless functions, and finally Kubernetes with TensorFlow Serving.

Stack

Python · scikit-learn · TensorFlow · Flask · Docker · Kubernetes, the same tools the course uses, kept close to how each module teaches them.

Takeaway

The part that stuck with me wasn't any single model. It was seeing a trained notebook become a service. Going from regression to deep learning and then deploying that work with Flask, Docker, and Kubernetes is what turned the coursework into a working understanding of how an ML project actually ships.