Description |
Data Analytics has become ubiquitous over the past decades, as tools based on Machine Learning continue producing innovative results in increasingly more fields. This course provides the foundations to autonomously understand and navigate the topic, with a practice-first approach that yet remains well-grounded into the theory. Students are confronted with weekly assignments, which begin without assumptions on prior knowledge, but soon explore in depth the Python programming language, data handling and pre-processing, visualization and plotting, all the way into the foundation of Machine Learning proper. The skills and confidence acquired are immediately applicable to real-world challenges, both in an industrial, product-first environment, and towards modern challenges in research-oriented academic laboratories. The course is entirely in English, and available at all time on Moodle. The students are encouraged to follow the lectures at their own pace, and start at any time, with the Moodle forums and frontal lectures providing extra support. Main topics: - Python programming: language fundamentals, Jupyter notebooks, object-oriented programming, advanced functionalities, core libraries, debugging.
- Handling data: data sources, defects and mitigation, access and formats, visualization.
- Problem analysis: experiment design, data selection, normalization, feature selection, fundamental statistics.
- Machine Learning: classification, regression, clustering, imputation, dimensionality reduction, with a few selected algorithms on each topic.
- Study of modern libraries and tools: numpy, jupyter, pandas, seaborn/matplotlib, scikit-learn, keras/tensorflow.
Code : UE-I09.00013 |