Franco Scarselli

Università degli Studi di Siena | UNISI · Department of Information Engineering and Mathematical

This paper aims to study the potentialities of incorporating recursive layers into Generative Adversarial Networks (GANs). Drawing inspiration from biological systems, in which feedback connections are prevalent, different studies investigated their impact on artificial neural networks. These studies have shown that feedback connections improve per...

The SH-SY5Y neuroblastoma cell line is often used as an in vitro model of neuronal function and is widely applied to study the molecular events leading to Alzheimer’s disease (AD). Indeed, recently, basic research on SH-SY5Y cells has provided interesting insights for the discovery of new drugs and biomarkers for improved AD treatment and diagnosis...

Alzheimer’s Disease (AD) affects the quality of life of millions of people worldwide and represents one of the biggest challenges for the whole society. The SH-SY5Y neuroblastoma cell line is often used as an in vitro model of neuronal function and is widely applied to study the molecular events leading to AD. In the last few years, basic research...

Graph Neural Networks (GNNs) have become the leading paradigm for learning on (static) graph-structured data. However, many real-world systems are dynamic in nature, since the graph and node/edge attributes change over time. In recent years, GNN-based models for temporal graphs have emerged as a promising area of research to extend the capabilities...

Drug side effects (DSEs), or adverse drug reactions (ADRs), constitute an important health risk, given the approximately 197,000 annual DSE deaths in Europe alone. Therefore, during the drug development process, DSE detection is of utmost importance, and the occurrence of ADRs prevents many candidate molecules from going through clinical trials. Th...

Graph Neural Networks (GNNs) are a large class of relational models for graph processing. Recent theoretical studies on the expressive power of GNNs have focused on two issues. On the one hand, it has been proven that GNNs are as powerful as the Weisfeiler-Lehman test (1-WL) in their ability to distinguish graphs. Moreover, it has been shown that t...

Graph Neural Networks (GNNs) have known an important and fast development in the last decade, with many theoretical and practical innovations. Their main feature is the capability of processing graph structured data with minimal loss of structural information. This makes GNNs the ideal family of models for processing a wide variety of biological da...

Recently, deep learning methods have had a tremendous impact on computer vision applications, from image classification and semantic segmentation to object detection and face recognition. Nevertheless, the training of state-of-the-art neural network models is usually based on the availability of large sets of supervised data. Indeed, deep neural ne...

Deep learning is widely applied in bioinformatics and biomedical imaging, due to its ability to perform various clinical tasks automatically and accurately. In particular, the application of deep learning techniques for the automatic identification of glomeruli in histopathological kidney images can play a fundamental role, offering a valid decisio...

In several areas of science and engineering, data can be naturally represented in graph form, where nodes denote entities and edges stand for relationships between them. Graph Neural Networks (GNNs) are a well-known class of machine learning models for graph processing. In this paper, we present GNNkeras, a library, based on Keras, which allows the...

Drug Side--Effects (DSEs) have a high impact on public health, care system costs, and drug discovery processes. Predicting the probability of side--effects, before their occurrence, is fundamental to reduce this impact, in particular on drug discovery. Candidate molecules could be screened before undergoing clinical trials, reducing the costs in ti...

The increasing digitization and datification of all aspects of people’s daily life, and the consequent growth in the use of personal data, are increasingly challenging the current development and adoption of Machine Learning (ML). First, the sheer complexity and amount of data available in these applications strongly demands for ML algorithms that...

Drug Side-Effects (DSEs) have a high impact on public health, care system costs, and drug discovery processes. Predicting the probability of side-effects, before their occurrence, is fundamental to reduce this impact, in particular on drug discovery. Candidate molecules could be screened before undergoing clinical trials, reducing the costs in time...

In this article, we present a smart gravimetric system for the automatic security monitoring of the accesses to public places with some entrance ticket or pass required (e.g., railway stations, subway stations, museums, exhibitions). The main objective is to spot illegal behaviors, for example, two people trying to enter using only one ticket; simu...

Graph neural networks (GNN) have shown outstanding applications in fields where data is essentially represented as graphs (e.g., chemistry, biology, recommendation systems). In this vein, communication networks comprise many fundamental components that are naturally represented in a graph-structured manner (e.g., topology, routing, signal interfere...

Graph neural networks (GNN) have shown outstanding applications in many fields where data is fundamentally represented as graphs (e.g., chemistry, biology, recommendation systems). In this vein, communication networks comprise many fundamental components that are naturally represented in a graph-structured manner (e.g., topology, configurations, tr...

In this paper, we use Generative Adversarial Networks (GANs) to synthesize high-quality retinal images along with the corresponding semantic label-maps, instead of real images during training of a segmentation network. Different from other previous proposals, we employ a two-step approach: first, a progressively growing GAN is trained to generate t...

Multi-organ segmentation of X-ray images is of fundamental importance for computer aided diagnosis systems. However, the most advanced semantic segmentation methods rely on deep learning and require a huge amount of labeled images, which are rarely available due to both the high cost of human resources and the time required for labeling. In this pa...

The automatic segmentation of the aorta can be extremely useful in clinical practice, allowing the diagnosis of numerous pathologies to be sped up, such as aneurysms and dissections, and allowing rapid reconstructive surgery, essential in saving patients’ lives. In recent years, the success of Deep Learning (DL)-based decision support systems has i...

To overcome problems with the design of large networks, particularly with respect to the depth of the network, this paper presents a new model of convolutional neural networks (CNN) which features fully recursive convolutional layers (RCLs). An RCL is a generalization of the classic one-stage feedforward convolutional layer (CL) to fully direct fee...

Graph Neural Networks (GNNs) are a wide class of connectionist models for graph processing. They perform an iterative message passing operation on each node and its neighbors, to solve classification/ clustering tasks --- on some nodes or on the whole graph --- collecting all such messages, regardless of their order. Despite the differences among t...

Multi-organ segmentation of X-ray images is of fundamental importance for computer aided diagnosis systems. However, the most advanced semantic segmentation methods rely on deep learning and require a huge amount of labeled images, which are rarely available due to both the high cost of human resources and the time required for labeling. In this pa...

Drug Discovery is a fundamental and ever-evolving field of research. The design of new candidate molecules requires large amounts of time and money, and computational methods are being increasingly employed to cut these costs. Machine learning methods are ideal for the design of large amounts of potential new candidate molecules, which are naturall...

Many realworld domains involve information naturally represented by graphs, where nodes denote basic patterns while edges stand for relationships among them. The Graph Neural Network (GNN) is a machine learning model capable of directly managing graphstructured data. In the original framework, GNNs are inductively trained, adapting their parameters...

The generation of graph-structured data is an emerging problem in the field of deep learning. Various solutions have been proposed in the last few years, yet the exploration of this branch is still in an early phase. In sequential approaches, the construction of a graph is the result of a sequence of decisions, in which, at each step, a node or a g...

Multi–parametric prostate MRI (mpMRI) is a powerful tool to diagnose prostate cancer, though difficult to interpret even for experienced radiologists. A common radiological procedure is to compare a magnetic resonance image with similarly diagnosed cases. To assist the radiological image interpretation process, computerized Content–Based Image Retr...

Nuclear magnetic resonance (NMR) is a powerful and non–invasive diagnostic tool. However, NMR scanned images are often noisy due to patient motions or breathing. Although modern Computer Aided Diagnosis (CAD) systems, mainly based on Deep Learning (DL), together with expert radiologists, can obtain very accurate predictions, working with noisy data...

Background Timely recognition of malignant melanoma (MM) is challenging for dermatologists worldwide and represents the main determinant for mortality. Dermoscopic examination is influenced by dermatologists’ experience and fails to achieve adequate accuracy and reproducibility in discriminating atypical nevi (AN) from early melanomas (EM). Object...

Providing pixel–level supervisions for scene text segmentation is inherently difficult and costly, so that only few small datasets are available for this task. To face the scarcity of training data, previous approaches based on Convolutional Neural Networks (CNNs) rely on the use of a synthetic dataset for pre–training. However, synthetic data cann...

Binding site identification allows to determine the function-ality and the quaternary structure of protein–protein complexes. Variousapproaches to this problem have been proposed without reaching a viablesolution. Representing the interacting peptides as graphs, a correspon-dence graph describing their interaction can be built. Finding the maxi-mum...

This paper extends the fully recursive perceptron network (FRPN) model for vectorial inputs to include deep convolutional neural networks (CNNs) which can accept multi-dimensional inputs. A FRPN consists of a recursive layer, which, given a fixed input, iteratively computes an equilibrium state. The unfolding realized with this kind of iterative me...

Background and objectives: Deep learning models and specifically Convolutional Neural Networks (CNNs) are becoming the leading approach in many computer vision tasks, including medical image analysis. Nevertheless, the CNN training usually requires large sets of supervised data, which are often difficult and expensive to obtain in the medical fiel...

Providing pixel-level supervisions for scene text segmentation is inherently difficult and costly, so that only few small datasets are available for this task. To face the scarcity of training data, previous approaches based on Convolutional Neural Networks (CNNs) rely on the use of a synthetic dataset for pre-training. However, synthetic data cann...

The absence of large scale datasets with pixel–level supervisions is a significant obstacle for the training of deep convolutional networks for scene text segmentation. For this reason, synthetic data generation is normally employed to enlarge the training dataset. Nonetheless, synthetic data cannot reproduce the complexity and variability of natur...

Long-term sleep quality assessment is essential to diagnose sleep disorders and to continuously monitor the health status. However, traditional polysomnography techniques are not suitable for long-term monitoring, whereas, methods able to continuously monitor the sleep pattern in an unobtrusive way are needed. In this paper, we present a general pu...

In recent years, the use of deep learning is becoming increasingly popular in computer vision. However, the effective training of deep architectures usually relies on huge sets of annotated data. This is critical in the medical field where it is difficult and expensive to obtain annotated images. In this paper, we use Generative Adversarial Network...

In recent years, Deep Neural Networks (DNNs) have led to impressive results in a wide variety of machine learning tasks, typically relying on the existence of a huge amount of supervised data. However, in many applications (e.g., bio–medical image analysis), gathering large sets of labeled data can be very difficult and costly. Unsupervised domain...

The absence of large scale datasets with pixel-level supervisions is a significant obstacle for the training of deep convolutional networks for scene text segmentation. For this reason, synthetic data generation is normally employed to enlarge the training dataset. Nonetheless, synthetic data cannot reproduce the complexity and variability of natur...

During the last decade, deep learning and Convolutional Neural Networks (CNNs) have produced a devastating impact on computer vision, yielding exceptional results on a variety of problems, including analysis of medical images. Recently, these techniques have been extended to 3D images with the downside of a large increase in the computational load....

In this paper, we introduce a new method for the segmentation of bacterial colonies in solid agar plate images. The proposed approach comprises two contributions. First, a simple but nonetheless effective engine is devised to generate synthetic plate images. This engine overlays bacterial colony patches to existing background images, taking into ac...

The Vapnik-Chervonenkis dimension (VC-dim) characterizes the sample learning complexity of a classification model and it is often used as an indicator for the generalization capability of a learning method. The VC-dim has been studied on common feed-forward neural networks, but it has yet to be studied on Graph Neural Networks (GNNs) and Recursive...

Graphs are a natural choice to encode data in many real–world applications. In fact, a graph can describe a given pattern as a complex structure made up of parts (the nodes) and relationships between them (the edges). Despite their rich representational power, most of machine learning approaches cannot deal directly with inputs encoded by graphs. I...

Most of the leading Convolutional Neural Network (CNN) models for semantic segmentation exploit a large number of pixel–level annotations. Such a human based labeling requires a considerable effort that complicates the creation of large–scale datasets. In this paper, we propose a deep learning approach that uses bounding box annotations to train a...

Learning machines for pattern recognition, such as neural networks or support vector machines, are usually conceived to process real–valued vectors with predefined dimensionality even if, in many real–world applications, relevant information is inherently organized into entities and relationships between them. Instead, Graph Neural Networks (GNNs)...

Traditional supervised approaches realize an inductive learning process: A model is learnt from labeled examples, in order to predict the labels of unseen examples. On the other hand, transductive learning is less ambitious. It can be thought as a procedure to learn the labels on a training set, while, simultaneously, trying to guess the best label...

Sleep plays a vital role in good health and well-being throughout our life. Getting enough quality sleep at the right times can help protect mental and physical health, quality of life, and safety. Emerging wearable devices allow people to measure and keep track of sleep duration, patterns, and quality. Often, these approaches are intrusive and cha...

This paper proposes the combination of two state-of-the-art algorithms for processing graph input data, viz., the probabilistic mapping graph self organizing map, an unsupervised learning approach, and the graph neural network, a supervised learning approach. We organize these two algorithms in a cascade architecture containing a probabilistic mapp...

Relevance ranking consists in sorting a set of objects with respect to a given criterion. However, in personalized retrieval systems, the relevance criteria may usually vary among different users and may not be predefined. In this case, ranking algorithms that adapt their behavior from users' feedbacks must be devised. Two main approaches are propo...

In the last decade, connectionist models have been proposed that can process structured information directly. These methods, which are based on the use of graphs for the representation of the data and the relationships within the data, are particularly suitable for handling relational learning tasks. In this paper, two recently proposed architectur...

In this paper, we will apply, to the task of detecting web spam, a combination of the best of its breed algorithms for processing graph domain input data, namely, probability mapping graph self organizing maps and graph neural networks. The two connectionist models are organized into a layered architecture, consisting of a mixture of unsupervised a...

In this paper, we will apply a recently proposed connectionist model, namely, the Graph Neural Network, for processing the graph formed by considering each sentence in a document as a node and the relationship between two sentences as an edge. Using commonly accepted evaluation protocols, the ROGUE toolkit, the technique was applied to two text sum...

This paper introduces a novel approach for processing a general class of structured information, viz., a graph of graphs structure, in which each node of the graph can be described by another graph, and each node in this graph, in turn, can be described by yet another graph, up to a finite depth. This graph of graphs description may be used as an u...

Graph Neural Networks (GNNs) are a powerful tool for processing graphs, that represent a natural way to collect information coming from several areas of science and engineering - e.g. data mining, computer vision, molecular chemistry, molecular biology, pattern recognition -, where data are intrinsically organized in entities and relationships amon...

Principal component analysis is often thought of as a preprocessing step for blind source separation (BSS). Although second order methods have been proposed for BSS in the past, these approaches cannot be easily implemented by neural models. In this ...

In graphical pattern recognition, each data is represented as an arrangement of elements, that encodes both the properties of each element and the relations among them. Hence, patterns are modelled as labelled graphs where, in general, labels can be attached to both nodes and edges. Artificial neural networks able to process graphs are a powerful t...

In this paper, we will consider the approximation properties of a recently introduced neural network model called graph neural network (GNN), which can be used to process-structured data inputs, e.g., acyclic graphs, cyclic graphs, and directed or undirected graphs. This class of neural networks implements a function tau(G,n) is an element of IR(m)...

Many underlying relationships among data in several areas of science and engineering, e.g., computer vision, molecular chemistry, molecular biology, pattern recognition, and data mining, can be represented in terms of graphs. In this paper, we propose a new neural network model, called graph neural network (GNN) model, that extends existing neural...

This chapter provides an introduction and motivates the leading thread of the following ten chapters that were collected to present some of the most recent advances in neural processing models, concerning both the analysis of theoretical properties of novel neural architectures and the illustration of some real–world applications. Not pretending to...

This research book presents some of the most recent advances in neural information processing models including both theoretical concepts and practical applications. The contributions include: Advances in neural information processing paradigms - Self organizing structures - Unsupervised and supervised learning of graph domains - Neural grammar netw...

The problem of relevance ranking consists of sorting a set of objects with respect to a given criterion. Since users may prefer different relevance criteria, the ranking algorithms should be adaptable to the user needs. Two main approaches exist in literature for the task of learning to rank: 1) a score function, learned by examples, which evaluate...

XML is becoming increasingly popular as a language for representing many types of electronic documents. The consequence of the strict structural document description via XML is that a relatively new task in mining documents based on structural and/or content information has emerged. In this paper we investigate (1) the suitability of new unsupervis...

Term weighting systems are of crucial importance in Information Extraction and Information Retrieval applications. Common approaches to term weighting are based either on statistical or on natural language analysis. In this paper, we present a new algorithm that capitalizes from the advantages of both the strategies by adopting a machine learning a...

The Graph Neural Network is a relatively new machine learning method capable of encoding data as well as relationships between data elements. This paper applies the Graph Neural Network for the first time to a given learning task at an international competition on the classification of semi-structured documents. Within this setting, the Graph Neura...

In this article, we present a new approach to page ranking. The page rank of a collection of Web pages can be represented in a parameterized model, and the user requirements can be represented by a set of constraints. For a particular parameterization, namely, a linear combination of the page ranks produced by different forcing functions, and user...

Term weighting is a crucial task in many Information Retrieval applications. Common approaches are based either on statistical or on natural language analysis. In this paper, we present a new algorithm that capitalizes from the advantages of both the strategies. In the proposed method, the weights are computed by a parametric function, called Conte...

Recursive neural networks are a powerful tool for processing structured data. According to the recursive learning paradigm, the input information consists of directed positional acyclic graphs (DPAGs). In fact, recursive networks are fed following the partial order defined by the links of the graph. Unfortunately, the hypothesis of processing DPAGs...

Graph Neural Networks (GNNs) are a recently proposed connectionist model that extends previous neural methods to structured domains. GNNs can be applied on datasets that contain very general types of graphs and, under mild hypotheses, they have been proven to be universal approximators on graphical domains. Whereas most of the common approaches to...

Recursive neural networks (RNNs) and graph neural networks (GNNs) are two connectionist models that can directly process graphs. RNNs and GNNs exploit a similar processing framework, but they can be applied to different input domains. RNNs require the input graphs to be directed and acyclic, whereas GNNs can process any kind of graphs. The aim of t...