Class Data Science

This module aims to introduce the foundational techniques and methodologies for analysing data from the interdisciplinary perspective of a Data Scientist. At the beginning of this module, students learn about the diverse nature of data and the symbolic power of different data structures. This foundational understanding naturally leads to a second stage in which students learn how to interrogate and extract information from data and justify their choices. During this stage, students learn about statistical inference, hypothesis testing, Frequentist vs Bayesian approaches to data, correlation, and causation. Finally, in the third part of the module, students learn the basics of machine learning through the theory and practice of regression, classification, and dimensionality reduction methods.

Not applicable

Introduction to Data Science and roles in the field Data types and tabular data structures Descriptive statistics: central tendency and dispersion Univariate visualization: histograms, KDE, boxplots, outliers, skew Bivariate analysis: scatter and pair plots, linear and non-linear relationships Correlation: Pearson and Spearman, strength vs significance Categorical data: contingency tables, proportions, grouped summaries Exploratory Data Analysis (EDA) pipelines and data storytelling Machine learning essentials: supervised vs unsupervised, features and labels, data splitting, cross-validation, sklearn pipelines Linear regression: model fit, residuals, R², overfitting, interpretation Logistic regression: binary classification, sigmoid, log-odds, ROC, accuracy, F1 Clustering: k-means, elbow method, intro to hierarchical clustering, applications Model validation and evaluation Statistical thinking. Review and integration of concept

The module integrates innovative teaching methods such as vibe-coding — hands-on programming sessions where lecturer and students explore solutions in real time, fostering intuition and creativity in code. Assessment focuses on the process rather than solely the final product, valuing individual progress, experimentation, and problem-solving skills. A dialogical approach is emphasised, encouraging continuous interaction between lecturer and students and supporting the development of critical thinking through debate and shared reflection.

  Grus, J. (2019). Data science from scratch: first principles with python. O'Reilly Media.