1,721,017 research outputs found
Semi-supervised learning and applications : a game-theoretic perspective
Full text linkThis thesis is focused on semi-supervised learning (SSL) algorithms, a family of methods lying in between supervised and unsupervised learning. The main characteristic of SSL algorithms is that they exploit at the same time the structure of the data (their features) and the available labeling information to estimates the boundaries of the classes/clusters. For this reason, they are particularly suitable in a regime of scarcity of labeled data or in the cases whether the data annotation is expensive or time-consuming. Here, we will exploit a recent algorithm, rooted in the evolutionary game-theory, named “Graph Transduction Games”. The GTG algorithm explicitly models an SSL problem as a non-cooperative game where players represent the data and the strategies the possible labels. A player chooses a strategy and receives a payoff which is proportional to the choice of the other players and to their similarities. The game is iterated until all the players have chosen their best strategy, and no one has any incentive to change his/her choice. The final labeling is then a property that emerges by the players interactions, hence from the data. During the labeling process, the similarities between all the data are taken into account, creating a context in which similar points affect each other in deciding the final labeling assignment. The neighboring players (data), hence the context, help in situations in which intrinsic ambiguities in the data may lead to inconsistent class assignments. Within this thesis, the GTG algorithm and the context in which players are playing will be explored into applications like bioinformatics, natural language processing, computer vision, and pure machine learning problems
Two-dimensional impurity imaging in polar ice cores sparks new demand for automated image analysis
No full textDue to its micron-scale resolution and micro-destructiveness, laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) is especially suited for the analysis of the oldest and highly thinned sections of polar ice cores. State-of-the-art 2D elemental imaging by LA-ICP-MS has great potential for investigating the location of impurities on the ice sample surface and is crucial to avoid misinterpretation of ultra-fine resolution signals. The impurity imaging with LA-ICP-MS comprises several millions of laser shots fired over just a few square mm. This technique combines new chemical images with visual analysis and, in so doing raises new questions that may be answered through techniques in automated image analysis and computer vision. As an illustration of this new frontier, a selected set of key problems is presented, with first examples of how automated image analysis techniques can help solving them. This concerns the relationship between impurity localization and the grain boundary network as well as the paleoclimate significance of single line profiles along the main core axis. Ultimately, this demonstrates that it is the right time to spark an intensified exchange among the two scientific communities of computer vision and ice core science
Hierarchical glocal attention pooling for graph classification
No full textGraph pooling is an essential operation in Graph Neural Networks that reduces the size of an input graph while preserving its core structural properties. Existing pooling methods find a compressed representation considering the Global Topological Structures (e.g., cliques, stars, clusters) or Local information at node level (e.g., top-informative nodes). However, an effective graph pooling method does not hierarchically integrate both Global and Local graph properties. To this end, we propose a dual-fold Hierarchical Global Local Attention Pooling (HGLA-Pool) layer that exploits the aforementioned graph properties, generating more robust graph representations. Exhaustive experiments on nine publicly available graph classification benchmarks under standard metrics show that HGLA-Pool significantly outperforms eleven state-of-the-art models on seven datasets while being on par for the remaining two
RAT-CC: A Recurrent Autoencoder for Time-Series Compression and Classification
No full textThe growth of interconnected devices has led to an enormous volume of temporal data that requires specialized compression models for efficient storage. Besides this, most applications need to classify these data efficiently, and having to reconstruct the original data from the compressed representation to then classify them is not optimal. For this reason, we propose a Recurrent Autoencoder for Time-series Compression and Classification, termed RAT-CC, that allows to perform any classification task on the compressed representation without needing to reconstruct the original time-series data. RAT-CC leverages a Long Short-Term Memory (LSTM) recurrent autoencoder with a dual-loss function: the standard reconstruction loss to minimize reconstruction error; and an embedding loss to preserve relative distances in the compressed embedding space. This combined loss ensures that the learned embeddings remain meaningful for classification tasks while preserving the necessary information for reconstruction. We assess the compression and classification performance of RAT-CC on four datasets taken from different domains. RAT-CC is implemented in Keras and freely available at (https://github.com/ChJ4m3s/RAT-CC)
Leveraging artificial intelligence methods to map seagrass ecosystems in Italian Seas: Tackling human impact and climate change
Full text linkMarine coastal ecosystems (MCEs) are crucial for human health, playing a key role in climate change adaptation. However, MCEs are globally threatened by environmental and human pressures. This study applies Graph Neural Networks (GNNs) to model seagrass distribution in the Italian Seas using a dataset of 2244 spatial units with environmental, climatic, and anthropogenic factors harmonised at 4 km resolution. GNN models, including Graph Convolutional and Attention Networks, were benchmarked against traditional machine learning methods: Random Forest, Support Vector Machine, and Multi-Layer Perceptron. GNNs achieved comparable overall accuracy (91%) but delivered more spatially consistent predictions and higher F1-scores (0.89) for the minority class (seagrass presence). Sensitivity analysis identified climatic and human variables as key drivers of seagrass distribution. These insights support the implementation of blue Nature-based Solutions (NbS) to protect and restore seagrass habitats, aiding biodiversity conservation and climate change mitigation while guiding effective policymaking
Method For Determining An Optimal Naval Navigation Routes From Historical GNSS Data Of Naval Trajectories
No full textA determination method of determining optimal routes for navigating naval vehicles by analysis of historical GNSS data, which optimal routes consist of sequences of waypoints. The determination method comprises the steps of: reducing the data by using an extraction method to retain only significant positions of a trajectory while keeping information about speed and heading changes; evaluating trajectories similarity by using a distance measure; applying a clustering algorithm to verify whether different patterns of trajectories exist for ships navigating the same area and assess whether the clusters extracted automatically from the data reflect the actual known traffic flows inside a port area; and extracting the most representative trajectories from the clusters extracted and the corresponding waypoints obtained by a further clustering algorithm applied to the points of the trajectories belonging to the same cluster of trajectories
A gradient-boosted tree framework to model the ice thickness of the world's glaciers (IceBoost v1.1)
Full text linkKnowledge of glacier ice volumes is crucial for constraining future sea level potential, evaluating freshwater resources, and assessing impacts on societies, from regional to global. Motivated by the disparity in existing ice volume estimates, we present IceBoost, a global machine learning framework trained to predict ice thickness at arbitrary coordinates, thereby enabling the generation of spatially distributed thickness maps for individual glaciers. IceBoost is an ensemble of two gradient-boosted trees trained with 3.7 million globally available ice thickness measurements and an array of 39 numerical features. The model error is similar to those of existing models outside polar regions and is up to 30 %-40 % lower at high latitudes. Providing supervision by exposing the model to available glacier thickness measurements reduces the error by a factor of up to 2 to 3. A feature-ranking analysis reveals that geodetic data are the most informative variables, while ice velocity can improve the model performance by 6 % at high latitudes. A major feature of IceBoost is a capability to generalize outside the training domain, i.e. producing meaningful ice thickness maps in all regions of the world, including on the ice sheet peripheries
Context aware nonnegative matrix factorization clustering
No full textIn this article we propose a method to refine the clustering results obtained with the nonnegative matrix factorization (NMF) technique, imposing consistency constraints on the final labeling of the data. The research community focused its effort on the initialization and on the optimization part of this method, without paying attention to the final cluster assignments. We propose a game theoretic framework in which each object to be clustered is represented as a player, which has to choose its cluster membership. The information obtained with NMF is used to initialize the strategy space of the players and a weighted graph is used to model the interactions among the players. These interactions allow the players to choose a cluster which is coherent with the clusters chosen by similar players, a property which is not guaranteed by NMF, since it produces a soft clustering of the data. The results on common benchmarks show that our model is able to improve the performances of many NMF formulations
Hypergraph Isomorphism Using Association Hypergraphs
Full text linkAssociation graphs represent a classical tool to deal with the graph matching problem and recently the idea has been generalized to the case of hypergraphs. In this article, the potential of this approach is explored. The proposed framework uses a class of dynamical systems derived from the Baum-Eagon inequality in order to find the maximum (maximal) clique in the association hypergraph, that corresponds to the maximum (maximal) isomorphism between the hypergraphs to be matched. The proposed approach has extensively been tested with experiments on a large synthetic dataset, including hypergraphs of different cardinalities, order and connectivities. In particular the isomorphism version of the problem has been analyzed. The results obtained are impressive in terms of correctness, thus showing that, despite its simplicity, the Baum-Eagon dynamics has an outstanding capacity of finding globally optimal solutions and solving the hypergraph isomorphism problem.Association graphs represent a classical tool to deal with the graph matching problem and recently the idea has been generalized to the case of hypergraphs. In this article, the potential of this approach is explored. The proposed framework uses a class of dynamical systems derived from the Baum-Eagon inequality in order to find the maximum (maximal) clique in the association hypergraph, that corresponds to the maximum (maximal) isomorphism between the hypergraphs to be matched. The proposed approach has extensively been tested with experiments on a large synthetic dataset, including hypergraphs of different cardinalities, order and connectivities. In particular the isomorphism version of the problem has been analyzed. The results obtained are impressive in terms of correctness, thus showing that, despite its simplicity, the Baum-Eagon dynamics has an outstanding capacity of finding globally optimal solutions and solving the hypergraph isomorphism problem. (C) 2019 Elsevier B.V. All rights reserved
On association graph techniques for hypergraph matching
No full textAssociation graph techniques represent a classical approach to tackle the graph matching problem and recently the idea has been generalized to the case of hypergraphs. In this paper, we explore the potential of this approach in conjunction with a class of dynamical systems derived from the Baum-Eagon inequality. In particular, we focus on the pure isomorphism case and show, with extensive experiments on a large synthetic dataset, that despite its simplicity the Baum-Eagon dynamics does an excellent job at finding globally optimal solutions
- …
