Nueva ponencia en «V Congreso Nacional de Registradores 2019»

El pasado viernes 4 de octubre participe en una mesa redonda en el V Congreso Nacional de Registradores dentro del bloque de “AI & Data Analytics” donde se debatió las enormes posibilidades que aportan estas tecnologías a los servicios que prestan los Registros.

Mi aportación estuvo centrada en presentar las técnicas de aprendizaje automático (machine learning) como aliadas para las labores de los registradores, así como explicar el proceso que supone desarrollar una aplicación basada en machine learning, y las diferencisa sustanciales que hay con el proceso tradicional de desarrollo software.

En el blog del IIC puedes encontrar un resumen de la ponencia, y en la siguiente publicación un resumen de todo el congreso.

Inferring ECA-based rules for ambient intelligence using evolutionary feature extraction

Leila S. Shafti, Pablo A. Haya, Manuel García-Herranz, Eduardo Pérez

Journal of Ambient Intelligence and Smart Environments 5 (6), 563-587 (2013) [download] (JCR, IF 2012: 1.298, Q2)

One of the goals in Ambient Intelligence is to enable Intelligent Environments to take decisions based on the perceived context. In our previous work, we successfully explored how the inhabitants can communicate their own preferences with the environment using Event-Condition-Action (ECA) rules. The easiness of the communication language combined with an appropriate explanation mechanism gives trust to the Intelligent Environment actions. However, defining every preference, and maintaining them up-to-date can be cumbersome. Therefore, a complementary mechanism is required to learn from user behavior and adapt to small changes without being explicitly requested for. Inferring behaviors effectively from data collected from sensors in an Intelligent Environment is a challenging problem. The main issues include primitive representation of data, the necessity of a high number of sensors, and dealing with few training data collected in a short time. We present MFE3/GADR, an evolutionary constructive induction method to ease inferring inhabitants’ preferences from data collected from simple sensors. We show that this method detects successfully relevant sensors and constructs highly informative features that abstract relations among them. The constructed features, in addition to improving significantly the learning accuracy, break down and encapsulate the performance of inhabitants into decision trees that can easily be converted to ECA rules for further use in the Intelligent Environment. Comparing the empirical results show that our method can reduce a large set of complex ECA rules that represent the preferences to a smaller set of simple ECA rules.