No abstract

Motivated by the increasing availability of open and free Earth observation data through the Copernicus Sentinel missions, this study investigates the capacity of advanced computational models to automatically generate thematic layers, which in turn contribute to and facilitate the creation of land cover products. In concrete terms, we assess the practical and computational aspects of multi-class Sentinel-2 image segmentation based on a convolutional neural network and random forest approaches. The annotated learning set derives from data that is made available as result of the implementation of European Union’s INSPIRE Directive. Since this network of data sets remains incomplete in regard to some geographic areas, another objective of this work was to provide consistent and reproducible ways for machine-driven mapping of these gaps and a potential update of the existing ones. Finally, the performance analysis identifies the most important hyper-parameters, and provides hints on the models’ deployment and their transferability.

A brief history of AI 19 2.2 Recent developments in machine learning 20 2.3 Recent developments in social robots 21 2.4 Current challenges 22 2.5 Summary and conclusions 23 3 EU in the AI competitive global landscape 25 3.1 3.1.1 USA 25 3.2 Analysing the key features of the AI landscape 29 3.4 Technological capacity 34 3.5 Summary and conclusions 35 4 AI in the EU 37 4.1 Strategies and plans 37 4.1.1 The European Union 37 4.1.2 France 40 4.1.3 United Kingdom 41 4.1.4 Finland 43 4.2 Summary and conclusions 43 5 The AI ecosystem in China 45 5.3 Regional/local initiatives 49 5.5 Summary and conclusions 51 Table of contents Part 2: Multi-dimensional perspectives 6 Ethical and societal perspective 6.1 Introduction 6.2.1 Challenges at individual level 6.2.2 Challenges at societal level 6.3 Summary and conclusions 7 Legal perspective 7.3 The protection of AI inventions/creations by intellectual property rights 7.4 Regulatory approach 7.5 Summary and conclusions 8 Educational perspective 8.1 Introduction 8.2 AI impact on skills demand, learning and teaching 8.2.1 Direct AI impact on advanced skills demand 8.2.2 Impact of AI on learning 8.2.3 Impact of AI on teaching 8.3 AI skills and academic supply 8.4 Summary and conclusions 9 Economic perspective 9.2 Potential impact of AI on growth 9.3 Potential impact of AI on inequality 9.4 Summary and conclusions 10 Cybersecurity perspective 10.1 Background: AI and cybersecurity 10.2 Applications of AI in cybersecurity 10.4 Robustness of AI algorithms against malicious action 10.5 Summary and conclusions D D Table of contents 11 Computer processing and energy perspective 95 11.1 Introduction 95 11.2 Assessment of data centre (DC) energy consumption 96 demand for HPC. 96 11.3.1 CPU advancements: energy saving and parallelisation computing 97 solutions 98 11.4 Summary and conclusions 100 12 Data perspective 103 Summary and conclusions 113 10 The digital transformation of society has just begun: AI is central to this change and offers major opportunities to improve our lives.

In this article we present two novel multipurpose reservoir optimization algorithms named nested stochastic dynamic programming (nSDP) and nested reinforcement learning (nRL). Both algorithms are built as a combination of two algorithms; in the nSDP case it is (1) stochastic dynamic programming (SDP) and (2) nested optimal allocation algorithm (nOAA) and in the nRL case it is (1) reinforcement learning (RL) and (2) nOAA. The nOAA is implemented with linear and non-linear optimization. The main novel idea is to include a nOAA at each SDP and RL state transition, that decreases starting problem dimension and alleviates curse of dimensionality. Both nSDP and nRL can solve multi-objective optimization problems without significant computational expenses and algorithm complexity and can handle dense and irregular variable discretization. The two algorithms were coded in Java as a prototype application and on the Knezevo reservoir, located in the Republic of Macedonia. The nSDP and nRL optimal reservoir policies were compared with nested dynamic programming policies, and overall conclusion is that nRL is more powerful, but significantly more complex than nSDP.

No abstract