Observer-dependent Collective Behavior For Biologically-inspired Processing Models
The role of observers in computational learning models is rarely discussed but offers a unique perspective on the practice of analysis and holistic interpretation of data. In this paper, we present an observer-dependent perspective of data acquisition and preprocessing. Observers consist of observers that exhibit environmental relativism, biorealism, and collective behavior. Environmental realism is achieved by introducing a variety of viewpoints to a population of observers, while biorealism provides observers that have an innate component that resembles a genotype. Collectively, observers can enable emergent features in a given dataset, as well as contextual understanding. The individual and collective behavior of observers can serve as a preprocessing layer or pre-trained model for a variety of machines and deep learning models. In considering the collective behavior of our computational observers, we make a number of predictions about their behavior that may facilitate specific applications.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Beyond Episodic AI: Cognitive Field Networks for Biologically Inspired Persistent Cognition
Cognitive Field Theory (CFT) proposes that cognition arises from memory-dressed collective dynamics that generate a persistent macroscopic cognitive field. Here we develop a Cognitive Field Network (CFN), a recurrent Tra…
Understanding Electro-communication and Electro-sensing in Weakly Electric Fish using Multi-Agent Deep Reinforcement Learning
Weakly electric fish, like Gnathonemus petersii, use a remarkable electrical modality for active sensing and communication, but studying their rich electrosensing and electrocommunication behavior and associated neural a…
Multi-agent Reinforcement LearningLearning Low-dimensional Latent Dynamics from High-dimensional Observations: Non-asymptotics and Lower Bounds
In this paper, we focus on learning a linear time-invariant (LTI) model with low-dimensional latent variables but high-dimensional observations. We provide an algorithm that recovers the high-dimensional features, i.e. c…
Meta-LearningA model of figure ground organization incorporating local and global cues
Figure Ground Organization (FGO) -- inferring spatial depth ordering of objects in a visual scene -- involves determining which side of an occlusion boundary is figure (closer to the observer) and which is ground (furthe…
Role of connectivity anisotropies in the dynamics of cultured neuronal networks
Laboratory-grown, engineered living neuronal networks in vitro have emerged in the last years as an experimental technique to understand the collective behavior of neuronal assemblies in relation to their underlying conn…