<?xml version='1.0' encoding='utf-8'?>
<rss version="2.0"><channel><title>Blog | Estevez Alvarez</title><link>https://estevezalvarez.com/en/blog/</link><description>Optimization, machine learning and data engineering.</description><language>en</language><item><title>Why results change: capacity, data and the limits of a GNN</title><link>https://estevezalvarez.com/en/blog/por-que-cambian-resultados-gnn-optimizacion.html</link><guid>https://estevezalvarez.com/en/blog/por-que-cambian-resultados-gnn-optimizacion.html</guid><description>A mathematical reading of my experiments: how demand distributions, LP bounds and search reach help or hinder each method.</description></item><item><title>No silver bullet in optimization: what I learned by testing a GNN on AWS</title><link>https://estevezalvarez.com/en/blog/optimizacion-gnn-aws-sin-bala-de-plata.html</link><guid>https://estevezalvarez.com/en/blog/optimizacion-gnn-aws-sin-bala-de-plata.html</guid><description>From experimental setup to results: when restricting search helps, when learning adds no benefit, and why context changes the decision.</description></item><item><title>Capacitated facility location and machine learning: models, experiments and results</title><link>https://estevezalvarez.com/en/blog/aprendizaje-maquina-localizacion-analisis-resultados.html</link><guid>https://estevezalvarez.com/en/blog/aprendizaje-maquina-localizacion-analisis-resultados.html</guid><description>An experimental analysis of where machine learning helps logistics optimization, with results, classical comparisons and limits of the evidence.</description></item><item><title>Scenarios and SAA: deciding before demand is known</title><link>https://estevezalvarez.com/en/blog/localizacion-incertidumbre.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-incertidumbre.html</guid><description>Sensitivity, two-stage decisions, VSS, and computational uncertainty for evaluating a logistics network.</description></item><item><title>Candidate scoring: learning to filter without losing good decisions</title><link>https://estevezalvarez.com/en/blog/localizacion-scoring.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-scoring.html</guid><description>Logistic regression and boosting as optimizer filters: what they predict and how to measure the cost of excluding candidates.</description></item><item><title>K-means: finding municipal profiles without inventing natural categories</title><link>https://estevezalvarez.com/en/blog/localizacion-clustering.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-clustering.html</guid><description>Scaling, choosing k, silhouette, and stability for interpreting territorial clusters.</description></item><item><title>Population forecasting: trends, damping, and temporal testing</title><link>https://estevezalvarez.com/en/blog/localizacion-prevision-poblacion.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-prevision-poblacion.html</guid><description>How to project population without confusing census revisions with growth or future population with guaranteed orders.</description></item><item><title>Predicting demand: from a population baseline to Poisson and boosting</title><link>https://estevezalvarez.com/en/blog/localizacion-demanda-machine-learning.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-demanda-machine-learning.html</guid><description>How to define the target, validate geographically, and interpret errors without treating one marketplace as the whole market.</description></item><item><title>LNS: reorganizing part of a network to escape a local optimum</title><link>https://estevezalvarez.com/en/blog/localizacion-lns.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-lns.html</guid><description>Destruction, MILP repair, seeds, and convergence traces in large neighborhood search.</description></item><item><title>Greedy and local search: build quickly, then improve deliberately</title><link>https://estevezalvarez.com/en/blog/localizacion-voraz-busqueda-local.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-voraz-busqueda-local.html</guid><description>How immediate choices, neighborhood moves, and local optima work in capacitated assignment.</description></item><item><title>Linear relaxation: how much better could a solution become?</title><link>https://estevezalvarez.com/en/blog/localizacion-relajacion-lineal.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-relajacion-lineal.html</guid><description>Learn to interpret bounds, gaps, and capacity sensitivity without mistaking a fractional solution for an operational network.</description></item><item><title>MILP: turning a logistics decision into a verifiable model</title><link>https://estevezalvarez.com/en/blog/localizacion-modelo-milp.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-modelo-milp.html</guid><description>Variables, constraints, capacity, and costs: build the three-center example and verify its optimum.</description></item><item><title>From orders to a logistics network: an optimization and machine learning project</title><link>https://estevezalvarez.com/en/blog/localizacion-galpones-proyecto.html</link><guid>https://estevezalvarez.com/en/blog/localizacion-galpones-proyecto.html</guid><description>What I built, how the algorithms connect, and what the capacitated facility location experiments demonstrate.</description></item><item><title>Spatial data segmentation: concepts and applications</title><link>https://estevezalvarez.com/en/blog/segmentacion-datos-espaciales.html</link><guid>https://estevezalvarez.com/en/blog/segmentacion-datos-espaciales.html</guid><description>What spatial segmentation is, how it differs from classification, key algorithms and how to apply it in QGIS with Orfeo Toolbox and PostGIS using Python examples.</description></item><item><title>Python for data analysis: from CSV to insight</title><link>https://estevezalvarez.com/en/blog/python-analisis-datos.html</link><guid>https://estevezalvarez.com/en/blog/python-analisis-datos.html</guid><description>A practical guide to the Python data stack: loading, exploration, cleaning, analysis and visualization with pandas and matplotlib using real examples.</description></item><item><title>Microservices: scalable architecture and the importance of sound modeling</title><link>https://estevezalvarez.com/en/blog/microservicios-arquitectura-y-modelado.html</link><guid>https://estevezalvarez.com/en/blog/microservicios-arquitectura-y-modelado.html</guid><description>Why the data model is the most important decision in a microservices architecture, and how to define domain boundaries that do not break in production.</description></item><item><title>Kubernetes: container orchestration step by step</title><link>https://estevezalvarez.com/en/blog/kubernetes-orquestacion-contenedores.html</link><guid>https://estevezalvarez.com/en/blog/kubernetes-orquestacion-contenedores.html</guid><description>A practical Kubernetes guide: cluster architecture, key components, installation with Minikube and your first deployment from scratch.</description></item><item><title>Automatic image classification with GeoAI in QGIS</title><link>https://estevezalvarez.com/en/blog/geoai-qgis-clasificacion-automatica.html</link><guid>https://estevezalvarez.com/en/blog/geoai-qgis-clasificacion-automatica.html</guid><description>A step-by-step guide to automatically classifying satellite images with the GeoAI plugin in QGIS: installation, preparation, training and interpretation of results.</description></item><item><title>ETL with Python and geoinformation: production pipelines</title><link>https://estevezalvarez.com/en/blog/etl-python-datos-geoinformacion.html</link><guid>https://estevezalvarez.com/en/blog/etl-python-datos-geoinformacion.html</guid><description>How to design robust ETL pipelines with Python and PostgreSQL/PostGIS: layered architecture, explicit validation, traceability and real production examples.</description></item></channel></rss>
