ICE Emissions Estimation through Machine-learning techniques to support real-world R&D experiments

The emissions estimation, both during homologation and standard driving, is one of the new challenges that automotive industries have to face. The new European and American regulation will allow a lower and lower quantity of Carbon Monoxide emission and will require that all the vehicles have to be able to monitor their own pollutants production. Since numerical models are too computationally expensive and approximated, new solutions based on Machine Learning are replacing standard techniques. In this project we considered a real V12 Internal Combustion Engine to propose a novel approach pushing Random Forests to generate meaningful prediction also in extreme cases (extrapolation, very high frequency peaks, noisy instrumentation etc.). The present work proposes also a data preprocessing pipeline for strongly unbalanced datasets and a reinterpretation of the regression problem as a classification problem in a logarithmic quantized domain. Results have been evaluated for two different models representing a pure interpolation scenario (more standard) and an extrapolation scenario, to test the out of bounds robustness of the model. The employed metrics take into account different aspects which can affect the homologation procedure, so the final analysis will focus on combining all the specific performances together to obtain the overall conclusions.

SynBA: A contextualized Synonim-Based adversarial Attack for text classification

With the advent of high-performance computing devices, deep neural networks have gained a lot of popularity in solving many Natural Language Processing tasks. However, they are also vulnerable to adversarial attacks, which are able to modify the input text in order to mislead the target model. Adversarial attacks are a serious threat to the security of deep neural networks, and they can be used to craft adversarial examples that steer the model towards a wrong decision. In this dissertation, we propose SynBA, a novel contextualized synonym-based adversarial attack for text classification. SynBA is based on the idea of replacing words in the input text with their synonyms, which are selected according to the context of the sentence. We show that SynBA successfully generates adversarial examples that are able to fool the target model with a high success rate. We demonstrate three advantages of this proposed approach: (1) effective - it outperforms state-of-the-art attacks by semantic similarity and perturbation rate, (2) utility-preserving - it preserves semantic content, grammaticality, and correct types classified by humans, and (3) efficient - it performs attacks faster than other methods.

Machine Learning in Automatic Tabulation for Constraint Model Reformulation

Combinatorial decision and optimization problems belong to numerous applications, such as logistics and scheduling, and can be solved with various approaches. Boolean Satisfiability and Constraint Programming solvers are some of the most used ones and their performance is significantly influenced by the model chosen to represent a given problem. This has led to the study of model reformulation methods, one of which is tabulation, that consists in rewriting the expression of a constraint in terms of a table constraint. To apply it, one should identify which constraints can help and which can hinder the solving process. So far this has been performed by hand, for example in MiniZinc, or automatically with manually designed heuristics, in Savile Row. Though, it has been shown that the performances of these heuristics differ across problems and solvers, in some cases helping and in others hindering the solving procedure. However, recent works in the field of combinatorial optimization have shown that Machine Learning (ML) can be increasingly useful in the model reformulation steps. This thesis aims to design a ML approach to identify the instances for which Savile Row’s heuristics should be activated. Additionally, it is possible that the heuristics miss some good tabulation opportunities, so we perform an exploratory analysis for the creation of a ML classifier able to predict whether or not a constraint should be tabulated. The results reached towards the first goal show that a random forest classifier leads to an increase in the performances of 4 different solvers. The experimental results in the second task show that a ML approach could improve the performance of a solver for some problem classes.

Implementazione di algoritmi Tiny Machine Learning per il monitoraggio strutturale basato su emissioni acustiche

Il monitoraggio basato su emissioni acustiche (AE) guidate si è confermato tra le tecniche più affidabili nel campo del Non-Destructive Testing delle strutture planari, vista anche la sua semplicità implementativa, i bassi costi che lo caratterizzano, la non invasività e la possibilità di realizzare un sistema che agisca in maniera continuativa ed in tempo reale sfruttando reti di sensori permanentemente installati, senza la necessità di ispezioni periodiche. In tale contesto, è possibile sfruttare l’abilità dell’apprendimento automatico nell’individuazione dei pattern nascosti all’interno dei segnali grezzi registrati, ottenendo così informazioni utili ai fini dell’applicazione considerata. L’esecuzione on-edge dei modelli, ovvero sul punto di acquisizione, consente di superare le limitazioni imposte dal processamento centralizzato dei dati, con notevoli vantaggi in termini di consumo energetico, tempestività nella risposta ed integrità degli stessi. A questo scopo, si rivela però necessario sviluppare modelli compatibili con le stringenti risorse hardware dei dispositivi a basso costo tipicamente impiegati. In questo elaborato verranno prese in esame alcune tipologie di reti neurali artificiali per l’estrazione dell’istante di arrivo (ToA) di un’emissione acustica all’interno di una sequenza temporale, in particolare quelle convoluzionali (CNNs) ed una loro variante più recente, le CapsNet basate su rounting by agreement. L’individuazione dei ToA relativi al medesimo evento su segnali acquisiti in diverse posizioni spaziali consente infatti di localizzare la sorgente da cui esso è scaturito. Le dimensioni di questi modelli permettono di eseguire l’inferenza direttamente su edge-device. I risultati ottenuti confermano la maggiore robustezza delle tecniche di apprendimento profondo rispetto ai metodi statistici tradizionali nel far fronte a diverse tipologie di disturbo, in particolare negli scenari più critici dal punto di vista del rapporto segnale-rumore.

Structured output SVM for remote sensing image classification

In the recent years, kernel methods have revealed very powerful tools in many application domains in general and in remote sensing image classification in particular. The special characteristics of remote sensing images (high dimension, few labeled samples and different noise sources) are efficiently dealt with kernel machines. In this paper, we propose the use of structured output learning to improve remote sensing image classification based on kernels. Structured output learning is concerned with the design of machine learning algorithms that not only implement input-output mapping, but also take into account the relations between output labels, thus generalizing unstructured kernel methods. We analyze the framework and introduce it to the remote sensing community. Output similarity is here encoded into SVM classifiers by modifying the model loss function and the kernel function either independently or jointly. Experiments on a very high resolution (VHR) image classification problem shows promising results and opens a wide field of research with structured output kernel methods.

Advanced Kernel Methods For Remote Sensing Image Classification

Résumé Suite aux recentes avancées technologiques, les archives d'images digitales ont connu une croissance qualitative et quantitative sans précédent. Malgré les énormes possibilités qu'elles offrent, ces avancées posent de nouvelles questions quant au traitement des masses de données saisies. Cette question est à la base de cette Thèse: les problèmes de traitement d'information digitale à très haute résolution spatiale et/ou spectrale y sont considérés en recourant à des approches d'apprentissage statistique, les méthodes à noyau. Cette Thèse étudie des problèmes de classification d'images, c'est à dire de catégorisation de pixels en un nombre réduit de classes refletant les propriétés spectrales et contextuelles des objets qu'elles représentent. L'accent est mis sur l'efficience des algorithmes, ainsi que sur leur simplicité, de manière à augmenter leur potentiel d'implementation pour les utilisateurs. De plus, le défi de cette Thèse est de rester proche des problèmes concrets des utilisateurs d'images satellite sans pour autant perdre de vue l'intéret des méthodes proposées pour le milieu du machine learning dont elles sont issues. En ce sens, ce travail joue la carte de la transdisciplinarité en maintenant un lien fort entre les deux sciences dans tous les développements proposés. Quatre modèles sont proposés: le premier répond au problème de la haute dimensionalité et de la redondance des données par un modèle optimisant les performances en classification en s'adaptant aux particularités de l'image. Ceci est rendu possible par un système de ranking des variables (les bandes) qui est optimisé en même temps que le modèle de base: ce faisant, seules les variables importantes pour résoudre le problème sont utilisées par le classifieur. Le manque d'information étiquétée et l'incertitude quant à sa pertinence pour le problème sont à la source des deux modèles suivants, basés respectivement sur l'apprentissage actif et les méthodes semi-supervisées: le premier permet d'améliorer la qualité d'un ensemble d'entraînement par interaction directe entre l'utilisateur et la machine, alors que le deuxième utilise les pixels non étiquetés pour améliorer la description des données disponibles et la robustesse du modèle. Enfin, le dernier modèle proposé considère la question plus théorique de la structure entre les outputs: l'intègration de cette source d'information, jusqu'à présent jamais considérée en télédétection, ouvre des nouveaux défis de recherche. Advanced kernel methods for remote sensing image classification Devis Tuia Institut de Géomatique et d'Analyse du Risque September 2009 Abstract The technical developments in recent years have brought the quantity and quality of digital information to an unprecedented level, as enormous archives of satellite images are available to the users. However, even if these advances open more and more possibilities in the use of digital imagery, they also rise several problems of storage and treatment. The latter is considered in this Thesis: the processing of very high spatial and spectral resolution images is treated with approaches based on data-driven algorithms relying on kernel methods. In particular, the problem of image classification, i.e. the categorization of the image's pixels into a reduced number of classes reflecting spectral and contextual properties, is studied through the different models presented. The accent is put on algorithmic efficiency and the simplicity of the approaches proposed, to avoid too complex models that would not be used by users. The major challenge of the Thesis is to remain close to concrete remote sensing problems, without losing the methodological interest from the machine learning viewpoint: in this sense, this work aims at building a bridge between the machine learning and remote sensing communities and all the models proposed have been developed keeping in mind the need for such a synergy. Four models are proposed: first, an adaptive model learning the relevant image features has been proposed to solve the problem of high dimensionality and collinearity of the image features. This model provides automatically an accurate classifier and a ranking of the relevance of the single features. The scarcity and unreliability of labeled. information were the common root of the second and third models proposed: when confronted to such problems, the user can either construct the labeled set iteratively by direct interaction with the machine or use the unlabeled data to increase robustness and quality of the description of data. Both solutions have been explored resulting into two methodological contributions, based respectively on active learning and semisupervised learning. Finally, the more theoretical issue of structured outputs has been considered in the last model, which, by integrating outputs similarity into a model, opens new challenges and opportunities for remote sensing image processing.

Validating Co-Training Models for Web Image Classification

Co-training is a semi-supervised learning method that is designed to take advantage of the redundancy that is present when the object to be identified has multiple descriptions. Co-training is known to work well when the multiple descriptions are conditional independent given the class of the object. The presence of multiple descriptions of objects in the form of text, images, audio and video in multimedia applications appears to provide redundancy in the form that may be suitable for co-training. In this paper, we investigate the suitability of utilizing text and image data from the Web for co-training. We perform measurements to find indications of conditional independence in the texts and images obtained from the Web. Our measurements suggest that conditional independence is likely to be present in the data. Our experiments, within a relevance feedback framework to test whether a method that exploits the conditional independence outperforms methods that do not, also indicate that better performance can indeed be obtained by designing algorithms that exploit this form of the redundancy when it is present.

Prediction of Parkinson's disease tremor onset using artificial neural networks

In this paper we present the initial results using an artificial neural network to predict the onset of Parkinson's Disease tremors in a human subject. Data for the network was obtained from implanted deep brain electrodes. A tuned artificial neural network was shown to be able to identify the pattern of the onset tremor from these real time recordings.

Identification of Parkinson’s disease tremor onset using artificial neural networks

In this paper we consider the possibility of using an artificial neural network to accurately identify the onset of Parkinson’s Disease tremors in human subjects. Data for the network is obtained by means of deep brain implantation in the human brain. Results presented have been obtained from a practical study (i.e. real not simulated data) but should be regarded as initial trials to be discussed further. It can be seen that a tuned artificial neural network can act as an extremely effective predictor in these circumstances.

Prediction of Parkinson’s disease tremor onset using radial basis function neural networks

The possibility of using a radial basis function neural network (RBFNN) to accurately recognise and predict the onset of Parkinson’s disease tremors in human subjects is discussed in this paper. The data for training the RBFNN are obtained by means of deep brain electrodes implanted in a Parkinson disease patient’s brain. The effectiveness of a RBFNN is initially demonstrated by a real case study.

Optimising memory usage in n-tuple neural networks

The use of n-tuple or weightless neural networks as pattern recognition devices is well known (Aleksander and Stonham, 1979). They have some significant advantages over the more common and biologically plausible networks, such as multi-layer perceptrons; for example, n-tuple networks have been used for a variety of tasks, the most popular being real-time pattern recognition, and they can be implemented easily in hardware as they use standard random access memories. In operation, a series of images of an object are shown to the network, each being processed suitably and effectively stored in a memory called a discriminator. Then, when another image is shown to the system, it is processed in a similar manner and the system reports whether it recognises the image; is the image sufficiently similar to one already taught? If the system is to be able to recognise and discriminate between m-objects, then it must contain m-discriminators. This can require a great deal of memory. This paper describes various ways in which memory requirements can be reduced, including a novel method for multiple discriminator n-tuple networks used for pattern recognition. By using this method, the memory normally required to handle m-objects can be used to recognise and discriminate between 2^m — 2 objects.

Design of associative memories using cellular neural networks

Cellular neural networks (CNNs) have locally connected neurons. This characteristic makes CNNs adequate for hardware implementation and, consequently, for their employment on a variety of applications as real-time image processing and construction of efficient associative memories. Adjustments of CNN parameters is a complex problem involved in the configuration of CNN for associative memories. This paper reviews methods of associative memory design based on CNNs, and provides comparative performance analysis of these approaches.

Application of ARX neural networks to model the rate of penetration of petroleum wells drilling

Bit performance prediction has been a challenging problem for the petroleum industry. It is essential in cost reduction associated with well planning and drilling performance prediction, especially when rigs leasing rates tend to follow the projects-demand and barrel-price rises. A methodology to model and predict one of the drilling bit performance evaluator, the Rate of Penetration (ROP), is presented herein. As the parameters affecting the ROP are complex and their relationship not easily modeled, the application of a Neural Network is suggested. In the present work, a dynamic neural network, based on the Auto-Regressive with Extra Input Signals model, or ARX model, is used to approach the ROP modeling problem. The network was applied to a real oil offshore field data set, consisted of information from seven wells drilled with an equal-diameter bit.