Studio e Sperimentazione di Algoritmi di Reinforcement Learning Applicati a Video Game

Nella prima parte del mio lavoro viene presentato uno studio di una prima soluzione "from scratch" sviluppata da Andrew Karpathy. Seguono due miei miglioramenti: il primo modificando direttamente il codice della precedente soluzione e introducendo, come obbiettivo aggiuntivo per la rete nelle prime fasi di gioco, l'intercettazione della pallina da parte della racchetta, migliorando l'addestramento iniziale; il secondo é una mia personale implementazione utilizzando algoritmi più complessi, che sono allo stato dell'arte su giochi dell'Atari, e che portano un addestramento molto più veloce della rete.

On Recurrent and Deep Neural Networks

L'apprentissage profond est un domaine de recherche en forte croissance en apprentissage automatique qui est parvenu à des résultats impressionnants dans différentes tâches allant de la classification d'images à la parole, en passant par la modélisation du langage. Les réseaux de neurones récurrents, une sous-classe d'architecture profonde, s'avèrent particulièrement prometteurs. Les réseaux récurrents peuvent capter la structure temporelle dans les données. Ils ont potentiellement la capacité d'apprendre des corrélations entre des événements éloignés dans le temps et d'emmagasiner indéfiniment des informations dans leur mémoire interne. Dans ce travail, nous tentons d'abord de comprendre pourquoi la profondeur est utile. Similairement à d'autres travaux de la littérature, nos résultats démontrent que les modèles profonds peuvent être plus efficaces pour représenter certaines familles de fonctions comparativement aux modèles peu profonds. Contrairement à ces travaux, nous effectuons notre analyse théorique sur des réseaux profonds acycliques munis de fonctions d'activation linéaires par parties, puisque ce type de modèle est actuellement l'état de l'art dans différentes tâches de classification. La deuxième partie de cette thèse porte sur le processus d'apprentissage. Nous analysons quelques techniques d'optimisation proposées récemment, telles l'optimisation Hessian free, la descente de gradient naturel et la descente des sous-espaces de Krylov. Nous proposons le cadre théorique des méthodes à région de confiance généralisées et nous montrons que plusieurs de ces algorithmes développés récemment peuvent être vus dans cette perspective. Nous argumentons que certains membres de cette famille d'approches peuvent être mieux adaptés que d'autres à l'optimisation non convexe. La dernière partie de ce document se concentre sur les réseaux de neurones récurrents. Nous étudions d'abord le concept de mémoire et tentons de répondre aux questions suivantes: Les réseaux récurrents peuvent-ils démontrer une mémoire sans limite? Ce comportement peut-il être appris? Nous montrons que cela est possible si des indices sont fournis durant l'apprentissage. Ensuite, nous explorons deux problèmes spécifiques à l'entraînement des réseaux récurrents, à savoir la dissipation et l'explosion du gradient. Notre analyse se termine par une solution au problème d'explosion du gradient qui implique de borner la norme du gradient. Nous proposons également un terme de régularisation conçu spécifiquement pour réduire le problème de dissipation du gradient. Sur un ensemble de données synthétique, nous montrons empiriquement que ces mécanismes peuvent permettre aux réseaux récurrents d'apprendre de façon autonome à mémoriser des informations pour une période de temps indéfinie. Finalement, nous explorons la notion de profondeur dans les réseaux de neurones récurrents. Comparativement aux réseaux acycliques, la définition de profondeur dans les réseaux récurrents est souvent ambiguë. Nous proposons différentes façons d'ajouter de la profondeur dans les réseaux récurrents et nous évaluons empiriquement ces propositions.

Learning in neural networks and stochastic approximation methods with averaging

The problem of adjusting the weights (learning) in multilayer feedforward neural networks (NN) is known to be of a high importance when utilizing NN techniques in various practical applications. The learning procedure is to be performed as fast as possible and in a simple computational fashion, the two requirements which are usually not satisfied practically by the methods developed so far. Moreover, the presence of random inaccuracies are usually not taken into account. In view of these three issues, an alternative stochastic approximation approach discussed in the paper, seems to be very promising.

On-line learning in neural networks

On-line learning is one of the most powerful and commonly used techniques for training large layered networks and has been used successfully in many real-world applications. Traditional analytical methods have been recently complemented by ones from statistical physics and Bayesian statistics. This powerful combination of analytical methods provides more insight and deeper understanding of existing algorithms and leads to novel and principled proposals for their improvement. This book presents a coherent picture of the state-of-the-art in the theoretical analysis of on-line learning. An introduction relates the subject to other developments in neural networks and explains the overall picture. Surveys by leading experts in the field combine new and established material and enable non-experts to learn more about the techniques and methods used. This book, the first in the area, provides a comprehensive view of the subject and will be welcomed by mathematicians, scientists and engineers, whether in industry or academia.

Adverse drug reaction classification with deep neural networks

We study the problem of detecting sentences describing adverse drug reactions (ADRs) and frame the problem as binary classification. We investigate different neural network (NN) architectures for ADR classification. In particular, we propose two new neural network models, Convolutional Recurrent Neural Network (CRNN) by concatenating convolutional neural networks with recurrent neural networks, and Convolutional Neural Network with Attention (CNNA) by adding attention weights into convolutional neural networks. We evaluate various NN architectures on a Twitter dataset containing informal language and an Adverse Drug Effects (ADE) dataset constructed by sampling from MEDLINE case reports. Experimental results show that all the NN architectures outperform the traditional maximum entropy classifiers trained from n-grams with different weighting strategies considerably on both datasets. On the Twitter dataset, all the NN architectures perform similarly. But on the ADE dataset, CNN performs better than other more complex CNN variants. Nevertheless, CNNA allows the visualisation of attention weights of words when making classification decisions and hence is more appropriate for the extraction of word subsequences describing ADRs.

Deep convolutional neural networks for semantic video object segmentation

In this thesis, we propose to infer pixel-level labelling in video by utilising only object category information, exploiting the intrinsic structure of video data. Our motivation is the observation that image-level labels are much more easily to be acquired than pixel-level labels, and it is natural to find a link between the image level recognition and pixel level classification in video data, which would transfer learned recognition models from one domain to the other one. To this end, this thesis proposes two domain adaptation approaches to adapt the deep convolutional neural network (CNN) image recognition model trained from labelled image data to the target domain exploiting both semantic evidence learned from CNN, and the intrinsic structures of unlabelled video data. Our proposed approaches explicitly model and compensate for the domain adaptation from the source domain to the target domain which in turn underpins a robust semantic object segmentation method for natural videos. We demonstrate the superior performance of our methods by presenting extensive evaluations on challenging datasets comparing with the state-of-the-art methods.

End-to-end Deep Metric Learning con Vision-Language Model per il Fashion Image Captioning

L'image captioning è un task di machine learning che consiste nella generazione di una didascalia, o caption, che descriva le caratteristiche di un'immagine data in input. Questo può essere applicato, ad esempio, per descrivere in dettaglio i prodotti in vendita su un sito di e-commerce, migliorando l'accessibilità del sito web e permettendo un acquisto più consapevole ai clienti con difficoltà visive. La generazione di descrizioni accurate per gli articoli di moda online è importante non solo per migliorare le esperienze di acquisto dei clienti, ma anche per aumentare le vendite online. Oltre alla necessità di presentare correttamente gli attributi degli articoli, infatti, descrivere i propri prodotti con il giusto linguaggio può contribuire a catturare l'attenzione dei clienti. In questa tesi, ci poniamo l'obiettivo di sviluppare un sistema in grado di generare una caption che descriva in modo dettagliato l'immagine di un prodotto dell'industria della moda dato in input, sia esso un capo di vestiario o un qualche tipo di accessorio. A questo proposito, negli ultimi anni molti studi hanno proposto soluzioni basate su reti convoluzionali e LSTM. In questo progetto proponiamo invece un'architettura encoder-decoder, che utilizza il modello Vision Transformer per la codifica delle immagini e GPT-2 per la generazione dei testi. Studiamo inoltre come tecniche di deep metric learning applicate in end-to-end durante l'addestramento influenzino le metriche e la qualità delle caption generate dal nostro modello.

Pretraining Convolutional Neural Networks for Visual Recognition

Convolutional Neural Networks (CNN) have become the state-of-the-art methods on many large scale visual recognition tasks. For a lot of practical applications, CNN architectures have a restrictive requirement: A huge amount of labeled data are needed for training. The idea of generative pretraining is to obtain initial weights of the network by training the network in a completely unsupervised way and then fine-tune the weights for the task at hand using supervised learning. In this thesis, a general introduction to Deep Neural Networks and algorithms are given and these methods are applied to classification tasks of handwritten digits and natural images for developing unsupervised feature learning. The goal of this thesis is to find out if the effect of pretraining is damped by recent practical advances in optimization and regularization of CNN. The experimental results show that pretraining is still a substantial regularizer, however, not a necessary step in training Convolutional Neural Networks with rectified activations. On handwritten digits, the proposed pretraining model achieved a classification accuracy comparable to the state-of-the-art methods.

Mutual Information-based RBM Neural Networks

(Deep) neural networks are increasingly being used for various computer vision and pattern recognition tasks due to their strong ability to learn highly discriminative features. However, quantitative analysis of their classication ability and design philosophies are still nebulous. In this work, we use information theory to analyze the concatenated restricted Boltzmann machines (RBMs) and propose a mutual information-based RBM neural networks (MI-RBM). We develop a novel pretraining algorithm to maximize the mutual information between RBMs. Extensive experimental results on various classication tasks show the eectiveness of the proposed approach.

Predicting corticospinal excitability from pre-TMS EEG using convolutional neural networks

The amplitude of motor evoked potentials (MEPs) elicited by transcranial magnetic stimulation (TMS) of the primary motor cortex (M1) shows a large variability from trial to trial, although MEPs are evoked by the same repeated stimulus. A multitude of factors is believed to influence MEP amplitudes, such as cortical, spinal and motor excitability state. The goal of this work is to explore to which degree the variation in MEP amplitudes can be explained by the cortical state right before the stimulation. Specifically, we analyzed a dataset acquired on eleven healthy subjects comprising, for each subject, 840 single TMS pulses applied to the left M1 during acquisition of electroencephalography (EEG) and electromyography (EMG). An interpretable convolutional neural network, named SincEEGNet, was utilized to discriminate between low- and high-corticospinal excitability trials, defined according to the MEP amplitude, using in input the pre-TMS EEG. This data-driven approach enabled considering multiple brain locations and frequency bands without any a priori selection. Post-hoc interpretation techniques were adopted to enhance interpretation by identifying the more relevant EEG features for the classification. Results show that individualized classifiers successfully discriminated between low and high M1 excitability states in all participants. Outcomes of the interpretation methods suggest the importance of the electrodes situated over the TMS stimulation site, as well as the relevance of the temporal samples of the input EEG closer to the stimulation time. This novel decoding method allows causal investigation of the cortical excitability state, which may be relevant for personalizing and increasing the efficacy of therapeutic brain-state dependent brain stimulation (for example in patients affected by Parkinson’s disease).

Self-supervised Information Retrieval basato su deep metric learning

La tesi ha lo scopo di ricercare, esaminare ed implementare un sistema di Machine Learning, un Recommendation Systems per precisione, che permetta la racommandazione di documenti di natura giuridica, i quali sono già stati analizzati e categorizzati appropriatamente, in maniera ottimale, il cui scopo sarebbe quello di accompagnare un sistema già implementato di Information Retrieval, istanziato sopra una web application, che permette di ricercare i documenti giuridici appena menzionati.

Neural Networks

We present an overview of current research on artificial neural networks, emphasizing a statistical perspective. We view neural networks as parameterized graphs that make probabilistic assumptions about data, and view learning algorithms as methods for finding parameter values that look probable in the light of the data. We discuss basic issues in representation and learning, and treat some of the practical issues that arise in fitting networks to data. We also discuss links between neural networks and the general formalism of graphical models.

Automatic blood glucose classification for gestational diabetes with feature selection: decision trees vs neural networks

Automatic blood glucose classification may help specialists to provide a better interpretation of blood glucose data, downloaded directly from patients glucose meter and will contribute in the development of decision support systems for gestational diabetes. This paper presents an automatic blood glucose classifier for gestational diabetes that compares 6 different feature selection methods for two machine learning algorithms: neural networks and decision trees. Three searching algorithms, Greedy, Best First and Genetic, were combined with two different evaluators, CSF and Wrapper, for the feature selection. The study has been made with 6080 blood glucose measurements from 25 patients. Decision trees with a feature set selected with the Wrapper evaluator and the Best first search algorithm obtained the best accuracy: 95.92%.

Bayesian methods for neural networks

Bayesian techniques have been developed over many years in a range of different fields, but have only recently been applied to the problem of learning in neural networks. As well as providing a consistent framework for statistical pattern recognition, the Bayesian approach offers a number of practical advantages including a potential solution to the problem of over-fitting. This chapter aims to provide an introductory overview of the application of Bayesian methods to neural networks. It assumes the reader is familiar with standard feed-forward network models and how to train them using conventional techniques.