-
Time Dynamics of the Dutch Municipality Network
Authors:
Ivan Jokić,
Edgar van Boven,
Ioannis Manolopoulos,
Trivik Verma,
Gert Buiten,
Frank Pijpers,
Hans van Hooff,
Piet Van Mieghem
Abstract:
Based on data sets provided by Statistics Netherlands and the International Institute of Social History, we investigate the Dutch municipality merging process and the survivability of municipalities over the period 1830-2019. We examine the dynamics of the population and area per municipality and how their distributions evolved during the researched period. We apply a Network Science approach, whe…
▽ More
Based on data sets provided by Statistics Netherlands and the International Institute of Social History, we investigate the Dutch municipality merging process and the survivability of municipalities over the period 1830-2019. We examine the dynamics of the population and area per municipality and how their distributions evolved during the researched period. We apply a Network Science approach, where each node represents a municipality and the links represent the geographical interconnections between adjacent municipalities via roads, railways, bridges or tunnels which were available in each specific yearly network instance. Over the researched period, we find that the distributions of the logarithm of both the population and area size closely follow a normal and a logistic distribution respectively. The tails of the population distributions follow a power-law distribution, a phenomenon observed in community structures of many real-world networks. The dynamics of the area distribution are mainly determined by the merging process, while the population distribution is also driven by the natural population growth and migration across the municipality network. Finally, we propose a model of the Dutch Municipality Network that captures population increase, population migration between municipalities and the process of municipality merging. Our model allows for predictions of the population and area distributions over time.
△ Less
Submitted 10 November, 2023;
originally announced November 2023.
-
Cortico-cerebellar networks as decoupling neural interfaces
Authors:
Joseph Pemberton,
Ellen Boven,
Richard Apps,
Rui Ponte Costa
Abstract:
The brain solves the credit assignment problem remarkably well. For credit to be assigned across neural networks they must, in principle, wait for specific neural computations to finish. How the brain deals with this inherent locking problem has remained unclear. Deep learning methods suffer from similar locking constraints both on the forward and feedback phase. Recently, decoupled neural interfa…
▽ More
The brain solves the credit assignment problem remarkably well. For credit to be assigned across neural networks they must, in principle, wait for specific neural computations to finish. How the brain deals with this inherent locking problem has remained unclear. Deep learning methods suffer from similar locking constraints both on the forward and feedback phase. Recently, decoupled neural interfaces (DNIs) were introduced as a solution to the forward and feedback locking problems in deep networks. Here we propose that a specialised brain region, the cerebellum, helps the cerebral cortex solve similar locking problems akin to DNIs. To demonstrate the potential of this framework we introduce a systems-level model in which a recurrent cortical network receives online temporal feedback predictions from a cerebellar module. We test this cortico-cerebellar recurrent neural network (ccRNN) model on a number of sensorimotor (line and digit drawing) and cognitive tasks (pattern recognition and caption generation) that have been shown to be cerebellar-dependent. In all tasks, we observe that ccRNNs facilitates learning while reducing ataxia-like behaviours, consistent with classical experimental observations. Moreover, our model also explains recent behavioural and neuronal observations while making several testable predictions across multiple levels. Overall, our work offers a novel perspective on the cerebellum as a brain-wide decoupling machine for efficient credit assignment and opens a new avenue between deep learning and neuroscience.
△ Less
Submitted 28 October, 2021; v1 submitted 21 October, 2021;
originally announced October 2021.
-
Current State and Future Directions for Learning in Biological Recurrent Neural Networks: A Perspective Piece
Authors:
Luke Y. Prince,
Roy Henha Eyono,
Ellen Boven,
Arna Ghosh,
Joe Pemberton,
Franz Scherr,
Claudia Clopath,
Rui Ponte Costa,
Wolfgang Maass,
Blake A. Richards,
Cristina Savin,
Katharina Anna Wilmes
Abstract:
We provide a brief review of the common assumptions about biological learning with findings from experimental neuroscience and contrast them with the efficiency of gradient-based learning in recurrent neural networks. The key issues discussed in this review include: synaptic plasticity, neural circuits, theory-experiment divide, and objective functions. We conclude with recommendations for both th…
▽ More
We provide a brief review of the common assumptions about biological learning with findings from experimental neuroscience and contrast them with the efficiency of gradient-based learning in recurrent neural networks. The key issues discussed in this review include: synaptic plasticity, neural circuits, theory-experiment divide, and objective functions. We conclude with recommendations for both theoretical and experimental neuroscientists when designing new studies that could help bring clarity to these issues.
△ Less
Submitted 5 January, 2022; v1 submitted 11 May, 2021;
originally announced May 2021.
-
Differences in radio emission from similar M dwarfs in the binary system Ross 867-8
Authors:
L. H. Quiroga-Nuñez,
H. T. Intema,
J. R. Callingham,
J. Villadsen,
H. J. van Langevelde,
P. Jagannathan,
T. W. Shimwell,
E. P. Boven
Abstract:
Serendipitously, we have rediscovered radio emission from the binary system Ross 867 (M4.5V) and Ross 868 (M3.5V) while inspecting archival Giant Metrewave Radio Telescope (GMRT) observations. The binary system consists of two M-dwarf stars that share common characteristics such as spectral type, astrometric parameters, age and emission at infrared, optical and X-rays frequencies. The GMRT data at…
▽ More
Serendipitously, we have rediscovered radio emission from the binary system Ross 867 (M4.5V) and Ross 868 (M3.5V) while inspecting archival Giant Metrewave Radio Telescope (GMRT) observations. The binary system consists of two M-dwarf stars that share common characteristics such as spectral type, astrometric parameters, age and emission at infrared, optical and X-rays frequencies. The GMRT data at 610 MHz taken on July 2011 shows that the radio emission from Ross 867 is polarized and highly variable on hour time scales with a peak flux of 10.4 $\pm$ 0.7 mJy/beam. Additionally, after reviewing archival data from several observatories (VLA, GMRT, JVLA and LOFAR), we confirm that although both stars are likely coeval, only Ross 867 has been detected, while Ross 868 remains undetected at radio wavelengths. As they have a a large orbital separation, this binary stellar system provides a coeval laboratory to examine and constrain the stellar properties linked to radio activity in M dwarfs. We speculate that the observed difference in radio activity between the dwarfs could be due to vastly different magnetic field topologies or that Ross 867 has an intrinsically different dynamo.
△ Less
Submitted 4 December, 2019; v1 submitted 19 November, 2019;
originally announced November 2019.
