Universidad San Sebastián - Sede Bellavista, Santiago – Chile.
21 al 25 de Noviembre de 2022
Bienvenido a JCC2022
Responsible Artificial Intelligence
Ricardo Baeza Yates
In the first part we cover four current specific problems that motivate the needs of responsible AI: (1) discrimination (e.g., facial recognition, justice, sharing economy, language models); (2) phrenology (e.g., biometric based predictions); (3) stupid models (e.g., minimal adversarial AI) and (4) indiscriminate use of computing resources (e.g., large language models). These examples do have a personal bias but set the context for the second part where we address four challenges: (1) too many principles (e.g., principles vs. techniques), (2) cultural differences; (3) regulation and (4) our cognitive biases. We finish discussing what we can do to address these challenges in the near future to be able to develop responsible AI.
Generalized Planning: Languages, Solutions, and Decompositions
Generalized planning is the task of finding solutions for collections of planning problems rather that solutions for invidual problems. In the first part of the talk, I will present the model for generalized planning, and languagues to express such models, their solutions, and general decompositions of problems into subproblems. In the second part, we will see how solutions can be learned either by using combinatorial solvers or deep learning, and how decompositions can be learned using combinatorial solvers. Relations to RL/DRL will be also discussed as well as current challenges.
María José Escobar
Studying biological and sensory systems allows us to understand the computational principles used to extract information from the environment, inspire new signal/image processing techniques, and develop new technologies. Throughout this talk, we will first focus on retinal computations and the principles used to encode/decode features in the visual scene. Here, we will show an application in constrast equalization and retinal ganglion cells model to guide artificial agents. Secondly, we will talk about evolutive learning in robots focusing on gait-learning in legged robots. Lastly, we will also analyze the cortical circuit associated with decision making, the cortical-basal ganglia loop, to incorporate it into a robot controller, which could be a suitable solution to calibrate the balance between exploitation and exploration.
Procedural Content Generation for Games
Procedural Content Generation (PCG) is the automation of media production. This media can be anything usually produced by humans, such as poetry, paintings, music, architectural drawings, or film. PCG for games is the use of algorithms to build game content that a designer would typically create, such as textures, sound effects, maps, levels, characters, weapons, quests, or even game mechanics and rules.
This talk will focus on content directly related to game mechanics and player interaction, that is, on functional rather than cosmetic content. We will cover the different types of content that can be created, some of the algorithms involved, and how academic research and industry collaborate.
Methods for the analysis of Diffusion MRI tractography
Diffusion MRI (dMRI) is sensitive to the movement of water molecules in tissues. Thanks to dMRI tractography, it is possible to reconstruct the trajectories of the main white matter fiber bundles. Tractography datasets are composed of 3D polylines, also called streamlines or fibers. These datasets pose interesting computational challenges due to their large size, noise, and the complexity of the white matter structure. Numerous methods have been developed to analyze brain tractography and study brain connections. In many cases, fiber clustering is used to reduce the complexity of the information and to detect patterns between subjects. These methods are used to create atlases of reproducible connections, which are identified in new subjects using segmentation methods. Through the calculation and analysis of diffusion-derived indices for each tract, it is possible to study the alterations of the structural connectivity for different pathologies. In this talk, we will review the main methods for tractography analysis and some application examples.
Computational Network Neuroscience of Naturalistic Language Processing
Steven L. Small
Centro Nacional de Investigación en Inteligencia Artificial – CENIA
Pontificia Universidad Católica de Chile
22 November 2022: 12:30PM
Fecha límite de envío
25 de Septiembre de 2022
Notificación de aceptación
25 de Octubre de 2022
15 de Noviembre de 2022
Extended DeadlineDebido a la gran cantidad de solicitudes se ha decidido extender la fecha
Carlos Hernandez Ulloa
Monica Otero Ferreiro
TARIFAS DE INSCRIPCIÓN
|CATEGORIA||VALOR REBAJADO CLP *||VALOR NORMAL CLP|
|ESTUDIANTE PREGRADO Y POSTGRADO||$7.000||$8.000|
|ASISTENTE CON PUBLICACIÓN (IEEE - SCOPUS) [ SCCC - TICXED - AEI ]||$70.000||$78.000|
|ASISTENTE CON PUBLICACIÓN NO INDEXADA [ LMN - ET ]||$35.000||$39.000|
|ARTÍCULO ADICIONAL [ SCCC - TICXED - AEI - LMN - ET ]||$35.000||$39.000|
|DOCENTES Y PROFESIONALES||$42.000||$47.000|
* PUEDE OPTAR A VALOR REBAJADO SI REALIZA SU INSCRIPCIÓN HASTA EL 20 DE OCTUBRE DEL 2021.