Genetic algorithms and evolutionary computation in education

This book is organized around a simple premise: modern AI and data-intensive decision making reduce, again and again, to search and optimization. We must choose among competing possibilities under uncertainty and constraints – models and features, schedules and resources, rules and policies – and we must do so in spaces that are often jagged, partially discrete, and shaped by real-world requirements. The first part of the book lays the theoretical and methodological groundwork for approaching these problems with Evolutionary Computation, and in particular with Genetic Algorithms; the second part translates that groundwork into the educational domain, where learning processes, institutional rules, and ethical imperatives turn optimization from an abstract exercise into a concrete design discipline. The book offers a practice‑oriented synthesis of how Evolutionary Computation, especially Genetic Algorithms, can be used to design and govern complex educational systems under uncertainty. The book advances “optimization‑as‑design,” using multi‑objective methods that surface Pareto fronts so institutions can balance effectiveness, cost, equity, latency, and workload in transparent ways. The book contains references for a software companion package in R – geneticaR – with a R library and several example scripts. Part I develops the methodological foundations: representation and fitness design; variation and selection; constraint handling via penalties, repairs, and feasibility‑preserving encodings; and multi‑objective algorithms such as NSGA‑II, together with guidance on when to prefer gradient methods, when to wrap them with evolutionary search, and how to scale via parallel and surrogate evaluation. Part II translates these tools to Learning Analytics and Educational Data Mining, demonstrating curriculum sequencing, timetabling, resource allocation, early‑warning and intervention pipelines, collaborative group formation, and intelligent tutoring-often in prescriptive settings where the question is not only “who is at risk?” but “what should we do, when, and under which constraints?”. Throughout, the book treats ethics and governance as first‑class design objectives: fairness and privacy are encoded in fitness; documentation, auditability, and human‑in‑the‑loop oversight are emphasized; and recommendations are aligned with emerging regulatory frameworks (e.g., EU AI Act, GDPR).