Course
Digicomp Code PYTHML
Python Machine Learning Bootcamp («PYTHML»)
Course facts
- Understanding machine learning concepts and their classification in modern data strategies
- Understanding data analysis and pre-processing with Python – from raw data to ML-compatible features
- Applying central ML algorithms such as decision trees, random forests, regression and clustering
- Understanding model evaluation and optimisation with metrics, cross-validation and hyperparameter tuning
- Creating reproducible workflows with scikit-learn pipelines and best practices
- Implementing practical ML projects independently – from the idea to the functioning model
This compact 2-day intensive course provides a practical introduction to the world of machine learning with Python, the leading ecosystem for data science and AI.
What can you expect?
- Clear, structured structure: from the basics to advanced ML models
- Hands-on instead of theory: direct implementation with real data sets
- Current tools and best practices: Scikit-Learn, Pandas, Keras and much more
- Transferability to everyday working life: classification, forecasting models, segmentation
The aim is not only to understand, but also to use the data safely: you will learn how to prepare data, select suitable models, train and evaluate them and convert them into a production-oriented form.
1 Introduction to Machine Learning & Setup
Goal: Understand what ML is and prepare the technical setup
Contents:
- What is machine learning? (Supervised, Unsupervised, Reinforcement)
- Practical application examples (image recognition, text classification, recommendation systems)
- Overview of scikit-learn, Pandas, Numpy
- Setup: Jupyter Notebook, Python environment, scikit-learn, matplotlib, seaborn
- First data exploration with Pandas
2 Data preparation & feature engineering
Goal: Analyse, clean and prepare data
Contents:
- Exploratory data analysis (EDA) with Pandas, Seaborn
- Missing values, outlier handling
- Data types, encoding (one-hot, label)
- Feature scaling: StandardScaler, MinMaxScaler
- Introduction to train/test splits and cross-validation
3 Linear models & classification
Goal: Introduction to supervised learning with a focus on classification
Contents:
- Brief introduction to linear regression
- Logistic regression for classification
- Performance metrics: Accuracy, Precision, Recall, F1, Confusion Matrix
- Practice: Classification task with Titanic dataset
4 Decision Trees & Random Forest
Goal: Understanding of decision structures and ensemble methods
Contents:
- Decision trees: structure, depth, overfitting
- Random forests: bagging, feature importance
- Comparison with logistic regression
- Practice: prediction with a real data set (e.g. credit scoring)
5 Model comparison & hyperparameter tuning
Goal: Systematically improve models
Contents:
- Repetition: Bias-Variance-Tradeoff
- Cross-Validation (K-Fold)
- GridSearchCV and RandomisedSearchCV
- Comparison of multiple models: Logistic Regression, Decision Tree, Random Forest, SVM
- Practice: Competition setup with Scikit-Learn Pipeline
6 Unsupervised Learning & Clustering
Goal: Recognise data patterns without labels
Contents:
- K-Means Clustering
- Comparison with Hierarchical Clustering
- PCA (Principal Component Analysis) for dimension reduction
- Visualisation: 2D/3D scatterplots with labels
- Practice: Customer segmentation
7 Introduction to neural networks
Goal: Understanding the basics of deep learning
Contents:
- Perception unit (perceptron) & activation functions
- Feedforward networks with Keras / Tensorflow
- Avoiding overfitting: Dropout, regularisation
- Practice: Simple classification (e.g. MNIST)
8 Mini final project & best practices
Goal: Apply and reflect on what has been learnt
Contents:
- Mini-project: Independent ML task (e.g. spam classification, housing prices)
- Step-by-step: EDA, model selection, optimisation, evaluation
This course is characterised by a mix of guided exercises, practical examples, theory and hands-on scenarios that are adapted, extended or created by the participants. With active and passive learning methods, various tasks and lots of practice, this course creates a sound basic understanding of machine learning with Python.
Ideal for professionals from IT, data analytics, engineering, business intelligence and related fields who are familiar with Python and want to use machine learning specifically in projects or find a structured introduction.
We recommend the course «Introduction to Programming with Python (PYTHON)» as preparation: