-
Effect in the spectra of eigenvalues and dynamics of RNNs trained with Excitatory-Inhibitory constraint
Authors:
Cecilia Jarne,
Mariano Caruso
Abstract:
In order to comprehend and enhance models that describes various brain regions is important to study the dynamics of trained recurrent neural networks. Including Dales law in such models usually presents several challenges. However, this is an important aspect that allows computational models to better capture the characteristics of the brain. Here we present a framework to train networks using su…
▽ More
In order to comprehend and enhance models that describes various brain regions is important to study the dynamics of trained recurrent neural networks. Including Dales law in such models usually presents several challenges. However, this is an important aspect that allows computational models to better capture the characteristics of the brain. Here we present a framework to train networks using such constraint. Then we have used it to train them in simple decision making tasks. We characterized the eigenvalue distributions of the recurrent weight matrices of such networks. Interestingly, we discovered that the non-dominant eigenvalues of the recurrent weight matrix are distributed in a circle with a radius less than 1 for those whose initial condition before training was random normal and in a ring for those whose initial condition was random orthogonal. In both cases, the radius does not depend on the fraction of excitatory and inhibitory units nor the size of the network. Diminution of the radius, compared to networks trained without the constraint, has implications on the activity and dynamics that we discussed here.
△ Less
Submitted 9 January, 2023; v1 submitted 22 August, 2022;
originally announced August 2022.
-
An algorithm to represent inbreeding trees
Authors:
C. Jarne,
F A. Gómez Albarracín,
M. Caruso
Abstract:
Recent work has proven the existence of extreme inbreeding in a European ancestry sample taken from the contemporary UK population \cite{nature_01}. This result brings our attention again to a math problem related to inbreeding family trees and diversity. Groups with a finite number of individuals could give a variety of genetic relationships. { In previous works \cite{PhysRevE.92.052132, PhysRevE…
▽ More
Recent work has proven the existence of extreme inbreeding in a European ancestry sample taken from the contemporary UK population \cite{nature_01}. This result brings our attention again to a math problem related to inbreeding family trees and diversity. Groups with a finite number of individuals could give a variety of genetic relationships. { In previous works \cite{PhysRevE.92.052132, PhysRevE.90.022125, JARNE20191}, we have addressed the issue of building inbreeding trees for biparental reproduction using Markovian models. Here, we extend these studies by presenting an algorithm to generate and represent inbreeding trees with no overlapping generations. We explicitly assume a two-gender reproductory scheme, and we pay particular attention to the links between nodes. We show that even for a simple case with a relatively small number of nodes in the tree, there are a large number of possible ways to rearrange the links between generations. We present an open-source python code to generate the tree graph, the adjacency matrix, and the histogram of the links for each different tree representation. We show how this mapping reflects the difference between tree realizations, and how valuable information may be extracted upon inspection of these matrices. The algorithm includes a feature to average several tree realizations, obtain the connectivity distribution, and calculate the average and mean value. We used this feature to compare trees with a different number of generations and nodes. The code presented here, available in Git-Hub, may be easily modified to be applied to other areas of interest involving connections between individuals, extend the study to add more characteristics of the different nodes, etc.
△ Less
Submitted 1 February, 2021; v1 submitted 21 September, 2020;
originally announced September 2020.
-
Markovian simulation for ancestors trees
Authors:
C. Jarne,
M. Caruso
Abstract:
We present a computational model to reconstruct trees of ancestors for animals with sexual reproduction. Through a recursive algorithm combined with a random number generator, it is possible to reproduce the number of ancestors for each generation and use it to constraint the maximum number of the following generation. This new model allows to consider the reproductive preferences of particular sp…
▽ More
We present a computational model to reconstruct trees of ancestors for animals with sexual reproduction. Through a recursive algorithm combined with a random number generator, it is possible to reproduce the number of ancestors for each generation and use it to constraint the maximum number of the following generation. This new model allows to consider the reproductive preferences of particular species and combine several trees to simulate the behavior of a population. It is also possible to obtain a description analytically, considering the simulation as a theoretical stochastic process. Such process can be generalized in order to use an algorithm associated with it to simulate other similar processes of stochastic nature. The simulation is based in the theoretical model previously presented before.
△ Less
Submitted 17 July, 2018; v1 submitted 26 December, 2016;
originally announced December 2016.
-
Markov chain approach to the distribution of ancestors in species of biparental reproduction
Authors:
M. Caruso,
C. Jarne
Abstract:
We studied how to obtain a distribution for the number of ancestors in species of sexual reproduction. Present models concentrate on the estimation of distributions repetitions of ancestors in genealogical trees. It has been shown that is not possible to reconstruct the genealogical history of each species along all its generations by means of a geometric progression. This analysis demonstrates th…
▽ More
We studied how to obtain a distribution for the number of ancestors in species of sexual reproduction. Present models concentrate on the estimation of distributions repetitions of ancestors in genealogical trees. It has been shown that is not possible to reconstruct the genealogical history of each species along all its generations by means of a geometric progression. This analysis demonstrates that it is possible to rebuild the tree of progenitors by modeling the problem with a Markov chain. For each generation, the maximum number of possible ancestors is different. This brings huge problems for the resolution. We found a solution through a dilation of the sample space, although the distribution defined there takes smaller values respect to the initial problem. In order to correct the distribution for each generation, we introduced the invariance under gauge (local) group of dilations. These ideas can used to study the interaction of several processes and provide a new approach on the problem of the common ancestor. In the same direction, this model also provide some elements that can be used to improve models of animal reproduction.
△ Less
Submitted 3 October, 2014; v1 submitted 12 September, 2011;
originally announced September 2011.
