arXiv subjects

George Panagopoulos

Publications and source records attributed to George Panagopoulos.

24 records · Page 2Linked to original sources

arXiv2020

Multi-task Learning for Influence Estimation and Maximization

We address the problem of influence maximization when the social network is accompanied by diffusion cascades. In prior works, such information is used to compute influence probabilities, which is utilized by stochastic diffusion models in influence maximization. Motivated by the recent criticism on the effectiveness of diffusion models as well as the galloping advancements in influence learning, we propose IMINFECTOR (Influence Maximization with INFluencer vECTORs), a unified approach that uses representations learned from diffusion cascades to perform model-independent influence maximization that scales in real-world datasets. The first part of our methodology is a multi-task neural network that learns embeddings of nodes that initiate cascades (influencer vectors) and embeddings of nodes that participate in them (susceptible vectors). The norm of an influencer vector captures the ability of the node to create lengthy cascades and is used to estimate the expected influence spread and reduce the number of candidate seeds. In addition, the combination of influencer and susceptible vectors form the diffusion probabilities between nodes. These are used to reformulate the network as a bipartite graph and propose a greedy solution to influence maximization that retains the theoretical guarantees.We a pply our method in three sizable networks with diffusion cascades and evaluate it using cascades from future time steps. IMINFECTOR is able to scale in all of them and outperforms various competitive algorithms and metrics from the diverse landscape of influence maximization in terms of efficiency and seed set quality.

cs.SI↗

arXiv2020

A new branch of inflationary speed limits

We present a new mechanism for inflation which exhibits a speed limit on scalar motion, generating accelerated expansion even on a steep potential. This arises from explicitly integrating out the short modes of additional fields coupled to the inflaton $ϕ$ via a dimension six operator, yielding an expression for the effective action which includes a nontrivial (logarithmic) function of $(\partialϕ)^2$. The speed limit appears at the branch cut of this logarithm arising in a large flavor expansion, similarly to the square root branch cut in DBI inflation arising in a large color expansion. Finally, we describe observational constraints on the parameters of this model.

hep-th↗

arXiv2020

Performance in the Courtroom: Automated Processing and Visualization of Appeal Court Decisions in France

Artificial Intelligence techniques are already popular and important in the legal domain. We extract legal indicators from judicial judgment to decrease the asymmetry of information of the legal system and the access-to-justice gap. We use NLP methods to extract interesting entities/data from judgments to construct networks of lawyers and judgments. We propose metrics to rank lawyers based on their experience, wins/loss ratio and their importance in the network of lawyers. We also perform community detection in the network of judgments and propose metrics to represent the difficulty of cases capitalising on communities features.

cs.CL↗

arXiv2020

Primordial Black Holes from non-Gaussian tails

We develop a primordial black hole (PBH) production mechanism, deriving non-Gaussian tails from interacting quantum fields during early universe inflation. The multi-field potential landscape may contain relatively flat directions, as a result of energetically favorable adjustments of fields coupled to the inflaton. Such additional fields do not contribute to CMB fluctuations given a sufficient large-scale decay, related to a gap in the critical exponents computed using stochastic methods. Along such directions transverse to the inflaton, the field makes rare jumps to large values. Mixing with the inflaton leads to a substantial tail in the resulting probability distribution for the primordial perturbations. Incorporating a large number of flavors of fields ensures theoretical control of radiative corrections and a substantial abundance. This generates significant PBH production for a reasonable window of parameters, with the mass range determined by the time period of mixing and the inflationary Hubble scale. We analyze a particular model in detail, and then comment on a broader family of models in this class which suggests a mechanism for primordial seeds for early super-massive black holes in the universe. Along the way, we encounter an analytically tractable example of stochastic dynamics and provide some representative calculations of its correlations and probability distributions.

hep-th↗

arXiv2020

Multipoint correlators in multifield cosmology

Connected $N$-point amplitudes in quantum field theory are enhanced by a factor of $N!$ in appropriate regimes of kinematics and couplings, but the non-perturbative analysis of this for collider physics applications is subtle. We resolve this question for $N$-point correlation functions of cosmological perturbations in multifield inflation, and comment on its application to primordial non-Gaussianity. We find that they are calculably $N!$-enhanced using a simple model for the mixing of the field sectors which leads to a convolution of their probability distributions. This effect leads to model-dependent but interesting prospects for enhanced observational sensitivity.

hep-th↗

arXiv2018

A Review of Network Inference Techniques for Neural Activation Time Series

Studying neural connectivity is considered one of the most promising and challenging areas of modern neuroscience. The underpinnings of cognition are hidden in the way neurons interact with each other. However, our experimental methods of studying real neural connections at a microscopic level are still arduous and costly. An efficient alternative is to infer connectivity based on the neuronal activations using computational methods. A reliable method for network inference, would not only facilitate research of neural circuits without the need of laborious experiments but also reveal insights on the underlying mechanisms of the brain. In this work, we perform a review of methods for neural circuit inference given the activation time series of the neural population. Approaching it from machine learning perspective, we divide the methodologies into unsupervised and supervised learning. The methods are based on correlation metrics, probabilistic point processes, and neural networks. Furthermore, we add a data mining methodology inspired by influence estimation in social networks as a new supervised learning approach. For comparison, we use the small version of the Chalearn Connectomics competition, that is accompanied with ground truth connections between neurons. The experiments indicate that unsupervised learning methods perform better, however, supervised methods could surpass them given enough data and resources.

stat.ML↗