A new approach for estimating confidence in machine learning model predictions, specifically in regression tasks utilizing Earth observation data with a particular focus on mosquito abundance (MA) estimation, is proposed here. We leverage the Variational AutoEncoder architecture to derive a confidence metric by the latent space representations of Earth observation datasets. This methodology is pivotal in establishing a correlation between the Euclidean distance in latent representations and the absolute error in individual MA predictions. Our study focuses on Earth observation datasets from the Veneto region in Italy and the Upper Rhine Valley in Germany, considering areas significantly affected by mosquito populations. A key finding is a notable correlation of 0.46 between the absolute error of MA predictions and the proposed confidence metric. This correlation signifies a robust, new metric for quantifying the reliability and enhancing the trustworthiness of the AI/ML model predictions in the context of both Earth observation data analysis and mosquito abundance studies.

This study develops a Data-Driven Machine Learning (ML) pipeline to predict West Nile Virus (WNV) outbreak risk at NUTS31level for every month of the transmission period (May - October), utilizing Earth Observation (EO), Socioeconomical, and Entomological data. While most related works predict WNV annually, we assess the feasibility of a per-month prediction approach using the Binary Relevance (BR) method as a baseline and comparing it to the Classifier Chain (CC) approach. Testing on four regions in Greece namely Attica, Central Macedonia, Thrace, and Thessaly reveals that the CC method outperforms the BR method, achieving an F1Score of 0.55. Our results show that a fine temporal resolution is viable when individual predictions are treated as a sequence and linked using a chained classifier, demonstrating the effectiveness of the chained approach.

Edge Machine Learning (Edge ML), which shifts computational intelligence from cloud-based systems to edge devices, is attracting significant interest due to its evident benefits including reduced latency, enhanced data privacy, and decreased connectivity reliance. While these advantages are compelling, they introduce unique challenges absent in traditional cloudbased approaches. In this paper, we delve into the intricacies of Edge-based learning, examining the interdependencies among: (i) constrained data storage on Edge devices, (ii) limited computational power for training, and (iii) the number of learning classes. Through experiments conducted using our MAGNETO system, that focused on learning human activities via data collected from mobile sensors, we highlight these challenges and offer valuable perspectives on Edge ML.

As mobile traffic is dominated by content services (e.g., video), which typically use recommendation systems, the paradigm of network-friendly recommendations (NFR) has been proposed recently to boost the network performance by promoting content that can be efficiently delivered (e.g., cached at the edge). NFR increase the network performance, however, at the cost of being unfair towards certain contents when compared to the standard recommendations. This unfairness is a side effect of NFR that has not been studied in literature. Nevertheless, retaining fairness among contents is a key operational requirement for content providers. This paper is the first to study the fairness in NFR, and design fair-NFR. Specifically, we use a set of metrics that capture different notions of fairness, and study the unfairness created by existing NFR schemes. Our analysis reveals that NFR can be significantly unfair. We identify an inherent trade-off between the network gains achieved by NFR and the resulting unfairness, and derive bounds for this trade-off. We show that existing NFR schemes frequently operate far from the bounds, i.e., there is room for improvement. To this end, we formulate the design of Fair-NFR (i.e., NFR with fairness guarantees compared to the baseline recommendations) as a linear optimization problem. Our results show that the Fair-NFR can achieve high network gains (similar to non-fair-NFR) with little unfairness.

The rise of algorithmic decision making led to active researches on how to define and guarantee fairness, mostly focusing on one-shot decision making. In several important applications such as hiring, however, decisions are made in multiple stage with additional information at each stage. In such cases, fairness issues remain poorly understood. In this paper we study fairness in k-stage selection problems where additional features are observed at every stage. We first introduce two fairness notions, local (per stage) and global (final stage) fairness, that extend the classical fairness notions to the k-stage setting. We propose a simple model based on a probabilistic formulation and show that the locally and globally fair selections that maximize precision can be computed via a linear program. We then define the price of local fairness to measure the loss of precision induced by local constraints; and investigate theoretically and empirically this quantity. In particular, our experiments show that the price of local fairness is generally smaller when the sensitive attribute is observed at the first stage; but globally fair selections are more locally fair when the sensitive attribute is observed at the second stage – hence in both cases it is often possible to have a selection that has a small price of local fairness and is close to locally fair.

Modern cellular networks are becoming denser, less regularly planned, and increasingly heterogeneous, making performance analysis challenging. We develop a flexible and accurate model of such heterogeneous networks (HetNets) consisting of K tiers of randomly located base stations (BSs), with different densities, transmit powers, and radio access technologies (RATs). Our main goal is to understand the impact of flow level dynamics on such a system, assuming non-saturated users that randomly generate download requests (“flows”). We do so by deriving analytically the per flow delay, the load, the utilization and the congestion probability of BSs in different tiers. We base our analysis on stochastic geometry, to understand the impact of topological randomness and intraand inter-tier interaction, and queueing theory, to model the competition between concurrent flows within the same BS, for each RAT. This allows us to model the interference more realistically as a function of network load. We apply our model to the case of a 2-tier network based on LTE and Wi-Fi and study different user inter-tier association criteria, such as off-load, max-SINR association, and min-delay association. Our results provide some interesting qualitative and quantitative insights about the impact of these association policies and different traffic intensities.

In this paper, we develop a flexible and accurate analytical model of large networks with random base station (BS) placement, in order to understand the impact of key network parameters like BS density and load on the network performance. The main goal is to understand the flow level dynamics of such a system, assuming non-saturated users and studying the congestion statistics for BSs and the per flow delay. To achieve this, we base our analysis on two main tools: (a)stochastic geometry, to understand the impact of topological randomness and coverage maps and (b) queueing theory, to model the competition between concurrent flows within the same BS. Our model is then applied the populars Radio Access Technologies (RATs), such as LTE and WIFi. Our results provide some interesting qualitative and quantitative insights about the performance of those networks.

We describe a channel sounding measurement campaign for cellular broadband wireless communications with high speed trains that was carried out in the context of the project CORRIDOR. The campaign combines MIMO and carrier aggregation to achieve very high throughputs. We compare two different scenarios, the first one reflects a cellular deployment, where the base station is about 1km away from the railway line. The second scenario corresponds to a railway deployed network, where the base station is located directly next the railway line. We present the general parameters of the measurement campaign and some results of Power Delay Profiles and Doppler Spectra and their evolution over time. Finally we present a simple channel model that captures the main effects observed in the measurements.