Introduction to Business Analytics

Original title (in Spanish): Analítica para los Negocios. All materials are in Spanish.

All course materials — theory documents, guided practices, in-class workshops, and datasets — are published on this site as each week is released. The documents are built with Quarto and embed runnable R through webR, so students can execute every example directly in the browser, with no local installation required.

Course description

The course introduces the business-analytics process: how to turn raw data into information, knowledge and evidence that supports decisions in private and public organizations. It concentrates on five analytical tasks — summarizing, visualizing, clustering, classifying, and estimating regressions — implemented in R on business datasets. Generative AI runs through the whole semester as a working tool: students learn what LLMs are, how to write effective prompts, and how to direct AI coding agents (Claude Code, Cursor, VS Code) responsibly — the AI accelerates the work, but the student must understand and validate every step of the analysis.

How each week works

The course follows ICESI’s active-learning model, organized in four recurring moments:

1. Theory before class. Each content week publishes a mandatory theory resource (podcast or video plus a companion document with runnable R); definitions and concepts are not re-explained in class.
2. Weekly quiz. Preparation is verified with a concept quiz (on Intu) in every content week.
3. Guided practice. The class session operationalizes the theory step by step, making the analytical decisions explicit.
4. In-class workshop. Each session closes with an individual, graded application activity.

Learning outcomes

By the end of the course, students will be able to:

• Explain the core concepts of business analytics and the process that turns data into decisions, and identify the analytical task — summarize, visualize, cluster, classify, regress — that answers a given business question.
• Use R to import, explore, clean, transform, and describe data, and build clear visualizations with ggplot2 following good communication practices.
• Diagnose data-quality problems — missing values, duplicates, inconsistent categories, outliers — and carry an exploratory data analysis from raw file to analysis-ready dataset with documented decisions.
• Train and evaluate machine-learning models in R — CART, random forests, the Lasso, and k-means — using train/test splits and cross-validation, and comparing them against a baseline with the right metrics (confusion matrix, precision/recall, MAE/RMSE, silhouette).
• Interpret and communicate results to support an organizational decision.
• Use generative-AI tools — LLMs and coding agents — critically and responsibly: writing effective prompts, validating every output, and keeping the analysis traceable.

Schedule

Unit 1 — Course foundations

Week 1 — Presentación del curso y metodología. What business analytics is and the problems it solves; the week-by-week tour of the course; the learning dynamic (theory before class, quiz, guided application); evaluation rules and the final project; and the course’s AI policy, motivated by recent evidence on how AI use affects skill formation. Slides (PDF) · Reading: Shen & Tamkin (2026), How AI Impacts Skill Formation

Week 2 — Introducción a los LLMs. What a large language model is and why it matters for economics and business; capabilities, limits and hallucinations, and why every output must be verified; the anatomy of an effective prompt (context, task, output format, constraints) and iterating on responses; reusable skills for recurring tasks. Theory · Guided practice · Workshop

Week 3 — Fundamentos de R. The RStudio interface and the script-based workflow; R as a calculator; data types and special values (NA, NULL); objects and assignment; vectors, matrices and data frames (creation, indexing, filtering); functions, the help system, and installing vs. loading packages; the Environment, and projects with relative paths for reproducibility. Theory · Guided practice · Workshop

Week 4 — Manipulación y visualización de datos. The dplyr grammar verb by verb — select(), rename(), filter(), arrange(), mutate(), summarise(), group_by(), count() — for global and grouped KPIs, and ggplot2 as a layered system (data, aesthetics, geometries, labels, themes). The week’s product: a KPI table and the 2–3 charts that communicate it. Theory · Guided practice · Workshop · Dataset (cafeteria.csv)

Unit 2 — The analytics process and EDA

Week 5 — Proceso analítico y tipos de analítica. Business analytics as a process that turns data into actionable knowledge; translating business questions into analytical tasks (classification, prediction, segmentation, anomaly detection, optimization); the analytics workflow from question to decision; and the roles in an analytics team (data engineer, data analyst, data scientist, business analyst). The in-class exercise evaluates business questions and feeds directly into Deliverable 1 of the final project. Theory · Guided practice · Workshop

Week 6 — EDA: fuentes, limpieza y exploración. Data sources and data quality in real contexts; the diagnostic checklist — types, plausible ranges, missing values, duplicates, inconsistent categories, outliers; cleaning with transparent, documented decisions; and a minimal EDA. The product: a reproducible raw → clean → analysis-ready pipeline with a decision log. Theory · Guided practice · Workshop · Dataset (ferreteria_raw.csv)

Week 7 — Examen Parcial 1. First integrative written exam (20%), covering weeks 1–6. September 7–13.

Unit 3 — AI applied to data analysis

Week 8 — Agentes de código: Claude Code, Cursor y VS Code. AI assistants for data analysis and programming: what these tools are and their typical workflows; redoing with AI what was done “by hand” in the previous weeks (manipulation, cleaning, EDA, visualization); and responsible use — iterating prompts, reviewing generated code, and verifying results against expectations. Theory · Guided practice · Workshop · Starter project (proyecto_semana8.zip)

Week 9 — Presentación Avance del Proyecto (EDA). Deliverable 2 of the final project (5%): oral presentation of the exploratory analysis — given the business question, how it will be answered and what the EDA shows — with feedback from the professor and course assistants. September 21–27.

Unit 4 — Machine learning foundations

Week 10 — Fundamentos de Machine Learning. Machine learning as generalization; the standard pipeline — target and features, train/test split, training and hyperparameters, metrics against a baseline; cross-validation; and the classic failure modes: overfitting, data leakage, and badly chosen metrics. Complementary videos build the confusion matrix and MAE/RMSE so the metrics are in place before the supervised weeks. Theory · Guided practice · Workshop · Datasets: credito_taller10.csv · notas_taller10.csv · credito_evaluacion.csv · notas_evaluacion.csv

Unit 5 — Supervised learning

Week 11 — Clasificación: árboles y bosques. The full classification pipeline; reading a classifier through the confusion matrix — why accuracy misleads, and the precision/recall trade-off; CART (partitions, depth, pruning, interpretability vs. overfitting) and random forests (bagging and why ensembles generalize better); choosing a model by the cost of its errors. The workshop predicts customer churn. Theory · Guided practice · Workshop · Datasets: credito_clasificacion.csv · clientes_conectatel.csv

Week 12 — Regresión: árboles, bosques y la ruta Lasso. The regression pipeline with MAE/RMSE against a mean/median baseline; regularized linear regression (Lasso, with Ridge as an aside) and variable selection; regression trees and ensembles (random forest / XGBoost); and the trade-off between interpretability, performance and overfitting risk. The workshop prices apartments in Cali. Theory · Guided practice · Workshop · Datasets: notas_regresion.csv · apartamentos_cali.csv

Unit 6 — Unsupervised learning

Week 13 — Clustering: fundamentos y métricas. Clustering as segmentation without a target; k-means (similarity as distance, why scaling matters, centroids); choosing k with the elbow and silhouette methods plus the interpretability criterion; profiling and naming segments so they are actionable; and high-level alternatives (hierarchical clustering, DBSCAN). The guided practice segments songs with Spotify audio features; the workshop segments the members of a gym chain. Theory · Guided practice · Workshop · Datasets: spotify_canciones.csv · socios_califit.csv

Week 14 — Examen Parcial 2. Second integrative written exam (20%), covering weeks 8–13. October 26–31.

Week 15 — Simulacro de la presentación final. Each group presents a preliminary version of its final defense and receives feedback on substance and form. November 2–8.

Week 16 — Presentación del Proyecto Final. Final oral defense of the project (Deliverable 3, 20%). It counts as an exam: attendance is mandatory. November 9–14.

Evaluation

Component	Weight	Allowed use of AI
Weekly concept quizzes (every content week, on Intu)	25%	No AI
In-class workshops (one per session)	10%	Collaboration with AI — every decision must be explainable; the business conclusion is written without AI
Integrative written exam 1 (week 7)	20%	No AI
Integrative written exam 2 (week 14)	20%	No AI
Final project (three deliverables)	25%	Planning / collaboration with AI, depending on the deliverable

Quizzes and workshops have no make-ups; as a mitigation mechanism, the worst quiz grade and the worst workshop grade are dropped once per semester. AI use in graded work follows ICESI’s institutional five-level IAG scale — each activity states its maximum allowed level, and every use must be declared, verified, and traceable.

Final project

The course assigns no closed cases: each group receives a multi-variable business dataset, formulates its own business question, and answers it with one of the course’s analytical tasks — classification, regression, or segmentation. Three deliverables: (1) the business question (week 6 — a prerequisite for the later deliverables), (2) an oral progress presentation of the EDA (week 9, 5%), and (3) the final oral defense (week 16, 20%). There is no written document: the defense is the final product, the presenting group member may be chosen at random, and attendance follows the same rules as an exam.

Bibliography and resources

• Wickham, H. & Grolemund, G. (2017). R for Data Science. O’Reilly Media. — data manipulation, visualization, and workflow.
• James, G., Witten, D., Hastie, T. & Tibshirani, R. (2021). An Introduction to Statistical Learning: with Applications in R (2nd ed.). Springer. — ML foundations, classification, regularization, trees, unsupervised learning.
• Perkins, M., Roe, J. & Furze, L. (2025). Reimagining the Artificial Intelligence Assessment Scale (AIAS). Journal of University Teaching and Learning Practice, 22(7). — the framework behind the course’s AI-use levels.

Reference handouts from the previous edition of the course, still available as complementary reading (in Spanish): IA y Machine Learning · LLMs, Skills y Agentes · Prompts: cómo formular preguntas a una IA

---

Materials are released progressively as the semester advances.