MSc · CIC-IPN
Safe route generation
Research on safe route generation through analysis of crime data on mobile devices.
Research
GeoAI, cross-city transfer learning and responsible model generalization.
Doctoral research focused on when, how and under which limits predictive models trained in one geographic region can transfer to another with different data distributions and urban structure.
Researcher
Martín Heriberto Pérez Gómez
PhD researcher in Computer Science · CIC-IPN · Mexico
The current research program studies the transferability of spatio-temporal predictive models across heterogeneous cities, with emphasis on crime data, geospatial representation, distribution shift and reproducible validation.
The central question is not whether a model can be moved from one city to another, but what knowledge can transfer, when negative transfer appears, and which conditions support responsible generalization.
Current comparative work links Mexico City, Chicago and London and combines machine learning, GIS, temporal modeling, data harmonization and controlled cross-region evaluation.
Research trajectory
From safe routing to geographic transfer learning
PhD · CIC-IPN
Geographic transferability
Spatio-temporal prediction, cross-city transfer, distribution shift and limits of generalization across heterogeneous urban contexts.
International research stay
BISITE · University of Salamanca
Doctoral research stay focused on strengthening methodological evaluation and international collaboration.
Scientific output
Accepted work
Bibliographic claims are displayed only when an editorial or institutional state can be verified.
Accepted · Ready for Production
Transfer Learning Across Cities for Spatio-Temporal Crime Prediction: A Comparative Evaluation in Mexico City, Chicago, and London
Martin Heriberto Pérez Gómez · José Giovanni Guzmán Lugo · Miguel Jesús Torres Ruiz
Research and Innovation Forum 2026 Vol. 1 · Springer proceedings
The definitive citation and DOI will be linked when official Springer metadata becomes public.
Public research activity
Talks, conferences and interviews
IEEE CIC-IPN · 2026
Can AI learn crime patterns in one city and transfer them to another?
Public webinar record covering Mexico City, Chicago and London.
RII Forum 2026 · Athens
International research presentation
Presentation of the cross-city transfer learning work later accepted for Springer proceedings.
Public interview · 2022
Technology and social challenges
Interview published during the UNODC professional period.
Verified identity
Independent academic and professional sources
These external records help resolve the researcher identity independently of Majestic Tech.
ORCID
Persistent researcher identifier: 0009-0001-4032-4654
Open external source →SABER · CIC-IPN
Institutional record covering the completed MSc and enrolled PhD program.
Open external source →PIIG Laboratory · CIC-IPN
Institutional directory of the Intelligent Geospatial Information Processing Laboratory.
Open external source →IEEE VTools · CIC-IPN
Public record of the 2026 talk on cross-city transfer between Mexico City, Chicago and London.
Open external source →Thesis supervision · CIC-IPN
Institutional page connecting the MSc and doctoral research line to academic supervision.
Open external source →Research evidence designed to remain independently verifiable.
ORCID, institutional records, event evidence and future DOI metadata converge on one researcher identity.