Short Coding Course — South Africa
scikit-learn, NumPy, Pandas & OpenAI API
Move beyond writing scripts and start building intelligent systems. This 40-hour course takes you from the fundamentals of supervised and unsupervised learning through to deploying real AI-powered applications — using the same tools as South Africa's leading data science teams.
Types of machine learning (supervised, unsupervised, reinforcement), when to use each, and a complete end-to-end project walkthrough. Environment setup: Python, Jupyter, scikit-learn, pandas.
Arrays, DataFrames, data cleaning, handling missing values and outliers, feature encoding, scaling, and building robust scikit-learn Pipelines.
Linear regression, polynomial regression, and regularisation techniques (Ridge and Lasso). Understanding bias–variance tradeoff and choosing the right model complexity.
Logistic regression and decision boundaries, evaluation metrics (accuracy, precision, recall, F1, ROC-AUC), confusion matrices, and multi-class classification strategies.
Maximum-margin classification, the kernel trick, linear and non-linear SVMs — and when SVMs outperform simpler models.
Building and visualising decision trees, controlling overfitting with pruning, bagging, random forests and gradient boosting — with feature importance analysis.
Cross-validation, grid search and randomised search, learning curves, and strategies for avoiding data leakage — ensuring your model performs on real data, not just training data.
PCA for feature reduction and visualisation, K-Means and DBSCAN clustering — with practical use cases in customer segmentation and anomaly detection.
How perceptrons and activation functions work, building a simple neural network with Keras/TensorFlow, understanding when deep learning adds value over classical ML.
Calling the OpenAI and Anthropic APIs from Python, prompt engineering fundamentals, embeddings and retrieval basics. Build a complete AI-powered Python application as your capstone project.
Taught by Code College's senior development trainers — practitioners with real-world AI and data science experience, backed by over 20 years of professional developer training since 2004.
Every module includes working code. You train models, analyse results and debug predictions — building a portfolio of Jupyter notebooks you can show employers.
Attend in-person at our Woodmead campus or join 100% online from anywhere in South Africa.
Receive a Code College certificate on completion — recognised by employers across South Africa.
No. Matric-level mathematics — basic algebra and an understanding of averages — is sufficient. The course focuses on applying machine learning with Python libraries rather than deriving algorithms from scratch. No calculus or linear algebra is required.
You should be comfortable writing Python functions, loops, and working with files before attending. Our Python Fundamentals short course is the recommended prerequisite if you are new to Python.
The Python Data Analysis course focuses on data cleaning, exploration, and visualisation with Pandas and Matplotlib. This course builds on those skills and goes further — covering machine learning models with scikit-learn, model evaluation, neural network basics, and integrating real AI APIs (OpenAI and Anthropic).
Yes. The final module covers calling the OpenAI and Anthropic (Claude) APIs from Python, prompt engineering fundamentals, and using embeddings for retrieval. You will build a complete AI-powered Python application as your capstone project.
Yes. The course is available 100% online via live virtual classroom, or in-person at our Johannesburg (Woodmead) campus. Both options deliver the same instructor-led experience.
The course is 40 hours, delivered as a 5-day full-time intensive or on a part-time schedule for groups. The fee is R14,995 per delegate. Group and corporate rates are available on request — contact us to discuss.
Book the Python for AI & Machine Learning course — full-time (5 days) or part-time. Johannesburg and online.