Learn to represent documents as high-dimensional vectors, apply TF-IDF weighting, measure document similarity, and discover structure in text collections using geometric and statistical modeling—without linguistic or NLP assumptions.

A hands-on, notebook-driven project where you will explore, visualize, and cluster a dataset of superhero statistics using unsupervised learning techniques. You'll perform data cleaning, exploratory analysis, dimensionality reduction with PCA, cluster discovery with K-Means, and anomaly detection with Isolation Forest—all in a strictly linear, function-free Python workflow.

Explore the foundations of tokenization in modern language models through the lens of information theory and compression. Understand how subword vocabularies are constructed, how entropy shapes token distributions, and the theoretical limits of tokenization strategies.

A hands-on, step-by-step case study guiding learners through the process of cleaning and exploring a real-world e-commerce transactions dataset using pandas, matplotlib, and seaborn. The course emphasizes linear, notebook-style code execution without function definitions, focusing on practical data wrangling and visualization skills.

A hands-on project guiding learners through exploratory data analysis and visualization of a city's restaurant market using a real-world dataset. Students will clean data, analyze cuisine and price segments, compare cities, and uncover business opportunities in the HoReCa industry.