An investigation of two real time machine learning techniques that could enhance the adaptability of game AI agents

Developers strive to create innovative Artificial Intelligence (AI) behaviour in their games as a key selling point. Machine Learning is an area of AI that looks at how applications and agents can be programmed to learn their own behaviour without the need to manually design and implement each aspect of it. Machine learning methods have been utilised infrequently within games and are usually trained to learn offline before the game is released to the players. In order to investigate new ways AI could be applied innovatively to games it is wise to explore how machine learning methods could be utilised in real-time as the game is played, so as to allow AI agents to learn directly from the player or their environment. Two machine learning methods were implemented into a simple 2D Fighter test game to allow the agents to fully showcase their learned behaviour as the game is played. The methods chosen were: Q-Learning and an NGram based system. It was found that N-Grams and QLearning could significantly benefit game developers as they facilitate fast, realistic learning at run-time.

Hybrid Artificial Intelligence to Extract Patterns and Rules from Argumentative and Legal Texts

This Thesis is composed of a collection of works written in the period 2019-2022, whose aim is to find methodologies of Artificial Intelligence (AI) and Machine Learning to detect and classify patterns and rules in argumentative and legal texts. We define our approach “hybrid”, since we aimed at designing hybrid combinations of symbolic and sub-symbolic AI, involving both “top-down” structured knowledge and “bottom-up” data-driven knowledge. A first group of works is dedicated to the classification of argumentative patterns. Following the Waltonian model of argument and the related theory of Argumentation Schemes, these works focused on the detection of argumentative support and opposition, showing that argumentative evidences can be classified at fine-grained levels without resorting to highly engineered features. To show this, our methods involved not only traditional approaches such as TFIDF, but also some novel methods based on Tree Kernel algorithms. After the encouraging results of this first phase, we explored the use of a some emerging methodologies promoted by actors like Google, which have deeply changed NLP since 2018-19 — i.e., Transfer Learning and language models. These new methodologies markedly improved our previous results, providing us with best-performing NLP tools. Using Transfer Learning, we also performed a Sequence Labelling task to recognize the exact span of argumentative components (i.e., claims and premises), thus connecting portions of natural language to portions of arguments (i.e., to the logical-inferential dimension). The last part of our work was finally dedicated to the employment of Transfer Learning methods for the detection of rules and deontic modalities. In this case, we explored a hybrid approach which combines structured knowledge coming from two LegalXML formats (i.e., Akoma Ntoso and LegalRuleML) with sub-symbolic knowledge coming from pre-trained (and then fine-tuned) neural architectures.

Development of advanced methods for the simulation of the reacting mixture formation in internal combustion engines with the use of machine learning algorithms

Besides increasing the share of electric and hybrid vehicles, in order to comply with more stringent environmental protection limitations, in the mid-term the auto industry must improve the efficiency of the internal combustion engine and the well to wheel efficiency of the employed fuel. To achieve this target, a deeper knowledge of the phenomena that influence the mixture formation and the chemical reactions involving new synthetic fuel components is mandatory, but complex and time intensive to perform purely by experimentation. Therefore, numerical simulations play an important role in this development process, but their use can be effective only if they can be considered accurate enough to capture these variations. The most relevant models necessary for the simulation of the reacting mixture formation and successive chemical reactions have been investigated in the present work, with a critical approach, in order to provide instruments to define the most suitable approaches also in the industrial context, which is limited by time constraints and budget evaluations. To overcome these limitations, new methodologies have been developed to conjugate detailed and simplified modelling techniques for the phenomena involving chemical reactions and mixture formation in non-traditional conditions (e.g. water injection, biofuels etc.). Thanks to the large use of machine learning and deep learning algorithms, several applications have been revised or implemented, with the target of reducing the computing time of some traditional tasks by orders of magnitude. Finally, a complete workflow leveraging these new models has been defined and used for evaluating the effects of different surrogate formulations of the same experimental fuel on a proof-of-concept GDI engine model.

Thermal and air quality modeling of an electric vehicle cabin with low-cost sensors and reinforcement learning

The design process of any electric vehicle system has to be oriented towards the best energy efficiency, together with the constraint of maintaining comfort in the vehicle cabin. Main aim of this study is to research the best thermal management solution in terms of HVAC efficiency without compromising occupant’s comfort and internal air quality. An Arduino controlled Low Cost System of Sensors was developed and compared against reference instrumentation (average R-squared of 0.92) and then used to characterise the vehicle cabin in real parking and driving conditions trials. Data on the energy use of the HVAC was retrieved from the car On-Board Diagnostic port. Energy savings using recirculation can reach 30 %, but pollutants concentration in the cabin builds up in this operating mode. Moreover, the temperature profile appeared strongly nonuniform with air temperature differences up to 10° C. Optimisation methods often require a high number of runs to find the optimal configuration of the system. Fast models proved to be beneficial for these task, while CFD-1D model are usually slower despite the higher level of detail provided. In this work, the collected dataset was used to train a fast ML model of both cabin and HVAC using linear regression. Average scaled RMSE over all trials is 0.4 %, while computation time is 0.0077 ms for each second of simulated time on a laptop computer. Finally, a reinforcement learning environment was built in OpenAI and Stable-Baselines3 using the built-in Proximal Policy Optimisation algorithm to update the policy and seek for the best compromise between comfort, air quality and energy reward terms. The learning curves show an oscillating behaviour overall, with only 2 experiments behaving as expected even if too slow. This result leaves large room for improvement, ranging from the reward function engineering to the expansion of the ML model.

Meaningful insights: explainability techniques for black-box models on tabular data

Artificial Intelligence (AI) and Machine Learning (ML) are novel data analysis techniques providing very accurate prediction results. They are widely adopted in a variety of industries to improve efficiency and decision-making, but they are also being used to develop intelligent systems. Their success grounds upon complex mathematical models, whose decisions and rationale are usually difficult to comprehend for human users to the point of being dubbed as black-boxes. This is particularly relevant in sensitive and highly regulated domains. To mitigate and possibly solve this issue, the Explainable AI (XAI) field became prominent in recent years. XAI consists of models and techniques to enable understanding of the intricated patterns discovered by black-box models. In this thesis, we consider model-agnostic XAI techniques, which can be applied to Tabular data, with a particular focus on the Credit Scoring domain. Special attention is dedicated to the LIME framework, for which we propose several modifications to the vanilla algorithm, in particular: a pair of complementary Stability Indices that accurately measure LIME stability, and the OptiLIME policy which helps the practitioner finding the proper balance among explanations' stability and reliability. We subsequently put forward GLEAMS a model-agnostic surrogate interpretable model which requires to be trained only once, while providing both Local and Global explanations of the black-box model. GLEAMS produces feature attributions and what-if scenarios, from both dataset and model perspective. Eventually, we argue that synthetic data are an emerging trend in AI, being more and more used to train complex models instead of original data. To be able to explain the outcomes of such models, we must guarantee that synthetic data are reliable enough to be able to translate their explanations to real-world individuals. To this end we propose DAISYnt, a suite of tests to measure synthetic tabular data quality and privacy.

Developing a radiomic based machine learning model to identify patients with resected non-small-cell lung cancer at high risk of relapse

Background There is a wide variation of recurrence risk of Non-small-cell lung cancer (NSCLC) within the same Tumor Node Metastasis (TNM) stage, suggesting that other parameters are involved in determining this probability. Radiomics allows extraction of quantitative information from images that can be used for clinical purposes. The primary objective of this study is to develop a radiomic prognostic model that predicts a 3 year disease free-survival (DFS) of resected Early Stage (ES) NSCLC patients. Material and Methods 56 pre-surgery non contrast Computed Tomography (CT) scans were retrieved from the PACS of our institution and anonymized. Then they were automatically segmented with an open access deep learning pipeline and reviewed by an experienced radiologist to obtain 3D masks of the NSCLC. Images and masks underwent to resampling normalization and discretization. From the masks hundreds Radiomic Features (RF) were extracted using Py-Radiomics. Hence, RF were reduced to select the most representative features. The remaining RF were used in combination with Clinical parameters to build a DFS prediction model using Leave-one-out cross-validation (LOOCV) with Random Forest. Results and Conclusion A poor agreement between the radiologist and the automatic segmentation algorithm (DICE score of 0.37) was found. Therefore, another experienced radiologist manually segmented the lesions and only stable and reproducible RF were kept. 50 RF demonstrated a high correlation with the DFS but only one was confirmed when clinicopathological covariates were added: Busyness a Neighbouring Gray Tone Difference Matrix (HR 9.610). 16 clinical variables (which comprised TNM) were used to build the LOOCV model demonstrating a higher Area Under the Curve (AUC) when RF were included in the analysis (0.67 vs 0.60) but the difference was not statistically significant (p=0,5147).

Machine learning per la previsione sui prezzi delle case

Negli ultimi anni, a causa della crescente tendenza verso i Big Data, l’apprendimento automatico è diventato un approccio di previsione fondamentale perché può prevedere i prezzi delle case in modo accurato in base agli attributi delle abitazioni. In questo elaborato, verranno messe in pratica alcune tecniche di machine learning con l’obiettivo di effettuare previsioni sui prezzi delle abitazioni. Ad esempio, si può pensare all’acquisto di una nuova casa, saranno tanti i fattori di cui si dovrà preoccuparsi, la posizione, i metri quadrati, l’inquinamento dell’aria, il numero di stanze, il numero dei bagni e così via. Tutti questi fattori possono influire in modo più o meno pesante sul prezzo di quell’abitazione. E’ proprio in casi come questi che può essere applicata l’intelligenza artificiale, nello specifico il machine learning, per riuscire a trovare un modello che approssimi nel miglior modo un prezzo, data una serie di caratteristiche. In questa tesi verrà dimostrato come è possibile utilizzare l’apprendimento automatico per effettuare delle stime il più preciso possibile dei prezzi delle case. La tesi è divisa in 5 capitoli, nel primo capitolo verranno introdotti i concetti di base su cui si basa l’elaborato e alcune spiegazioni dei singoli modelli. Nel secondo capitolo, invece, viene trattato l’ambiente di lavoro utilizzato, il linguaggio e le relative librerie utilizzate. Il terzo capitolo contiene un’analisi esplorativa sul dataset utilizzato e vengono effettuate delle operazioni per preparare i dati agli algoritmi che verranno applicati in seguito. Nel capitolo 4 vengono creati i diversi modelli ed effettuate le previsioni sui prezzi mentre nel capitolo 5 vengono analizzati i risultati ottenuti e riportate le conclusioni.

Machine Learning in ambito assicurativo Un caso di studio

Il Machine Learning si sta rivelando una tecnologia dalle incredibili potenzialità nei settori più disparati. Le diverse tecniche e gli algoritmi che vi fanno capo abilitano analisi dei dati molto più efficaci rispetto al passato. Anche l’industria assicurativa sta sperimentando l’adozione di soluzioni di Machine Learning e diverse sono le direzioni di innovamento che ne stanno conseguendo, dall’efficientamento dei processi interni all’offerta di prodotti rispondenti in maniera adattiva alle esigenze del cliente. Questo lavoro di tesi è stato realizzato durante un tirocinio presso Unisalute S.p.A., la prima assicurazione in ambito sanitario in Italia. La criticità intercettata è stata la sovrastima del capitale da destinare a riserva a fronte dell’impegno nei confronti dell’assicurato: questo capitale immobilizzato va a sottrarre risorse ad investimenti più proficui nel medio e lungo termine, per cui è di valore stimarlo appropriatamente. All'interno del settore IT di Unisalute, ho lavorato alla progettazione e implementazione di un modello di Machine Learning che riesca a prevedere se un sinistro appena preso in gestione sarà liquidato o meno. Dotare gli uffici impegnati nella determinazione del riservato di questa stima aggiuntiva basata sui dati, sarebbe di notevole supporto. La progettazione del modello di Machine Learning si è articolata in una Data Pipeline contenente le metodologie più efficienti con riferimento al preprocessamento e alla modellazione dei dati. L’implementazione ha visto Python come linguaggio di programmazione; il dataset, ottenuto a seguito di estrazioni e integrazioni a partire da diversi database Oracle, presenta una cardinalità di oltre 4 milioni di istanze caratterizzate da 32 variabili. A valle del tuning degli iperparamentri e dei vari addestramenti, si è raggiunta un’accuratezza dell’86% che, nel dominio di specie, è ritenuta più che soddisfacente e sono emersi contributi non noti alla liquidabilità dei sinistri.

Machine learning approach to the extended Hubbard model

The 1d extended Hubbard model with soft-shoulder potential has proved itself to be very difficult to study due its non solvability and to competition between terms of the Hamiltonian. Given this, we tried to investigate its phase diagram for filling n=2/5 and range of soft-shoulder potential r=2 by using Machine Learning techniques. That led to a rich phase diagram; calling U, V the parameters associated to the Hubbard potential and the soft-shoulder potential respectively, we found that for V<5 and U>3 the system is always in Tomonaga Luttinger Liquid phase, then becomes a Cluster Luttinger Liquid for 5and V), and finally undergoes a general crystallization or V>7, with a quasi-perfect crystal in the U<3V/2 and U>5 region. Finally we found that for U<5 and V>2-3 the system shall maintain the Cluster Luttinger Liquid structure, with a residual in-block single particle mobility.

Tecniche parallele di machine learning applicate a container

L’obiettivo di questa tesi `e l’estensione della conoscenza di un argomento già ampliamente conosciuto e ricercato. Questo lavoro focalizza la propria attenzione su una nicchia dell’ampio mondo della virtualizzazione, del machine learning e delle tecniche di apprendimento parallelo. Nella prima parte verranno spiegati alcuni concetti teorici chiave per la virtualizzazione, ponendo una maggior attenzione verso argomenti di maggior importanza per questo lavoro. La seconda parte si propone di illustrare, in modo teorico, le tecniche usate nelle fasi di training di reti neurali. La terza parte, attraverso una parte progettuale, analizza le diverse tecniche individuate applicandole ad un ambiente containerizzato.

tecniche di machine learning per la predizione del rating delle applicazioni all’interno di un play store

Il mio progetto di tesi ha come obiettivo quello di creare un modello in grado di predire il rating delle applicazioni presenti all’interno del Play Store, uno dei più grandi servizi di distribuzione digitale Android. A tale scopo ho utilizzato il linguaggio Python, che grazie alle sue librerie, alla sua semplicità e alla sua versatilità è certamen- te uno dei linguaggi più usati nel campo dell’intelligenza artificiale. Il punto di partenza del mio studio è stato il Dataset (Insieme di dati strutturati in forma relazionale) “Google Play Store Apps” reperibile su Kaggle al seguente indirizzo: https://www.kaggle.com/datasets/lava18/google-play-store-apps, contenente 10841 osservazioni e 13 attributi. Dopo una prima parte relativa al caricamen- to, alla visualizzazione e alla preparazione dei dati su cui lavorare, ho applica- to quattro di↵erenti tecniche di Machine Learning per la stima del rating delle applicazioni. In particolare, sono state utilizzate:https://www.kaggle.com/datasets/lava18/google-play-store-apps, contenente 10841 osservazioni e 13 attributi. Dopo una prima parte relativa al caricamento, alla visualizzazione e alla preparazione dei dati su cui lavorare, ho applicato quattro differenti tecniche di Machine Learning per la stima del rating delle applicazioni: Ridje, Regressione Lineare, Random Forest e SVR. Tali algoritmi sono stati applicati attuando due tipi diversi di trasformazioni (Label Encoding e One Hot Encoding) sulla variabile ‘Category’, con lo scopo di analizzare come le suddette trasformazioni riescano a influire sulla bontà del modello. Ho confrontato poi l’errore quadratico medio (MSE), l’errore medio as- soluto (MAE) e l’errore mediano assoluto (MdAE) con il fine di capire quale sia l’algoritmo più efficiente.

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.

Sviluppo di un sistema di climatizzazione automatico distribuito con l'utilizzo di tecniche di machine learning

Nella sede dell’azienda ospitante Alexide, si è ravvisata la mancanza di un sistema di controllo automatico da remoto dell’intero impianto di climatizzazione HVAC (Heating, Ventilation and Air Conditioning) utilizzato, e la soluzione migliore è risultata quella di attuare un processo di trasformazione della struttura in uno smart building. Ho quindi eseguito questa procedura di trasformazione digitale progettando e sviluppando un sistema distribuito in grado di gestire una serie di dati provenienti in tempo reale da sensori ambientali. L’architettura del sistema progettato è stata sviluppata in C# su ambiente dotNET, dove sono stati collezionati i dati necessari per il funzionamento del modello di predizione. Nella fattispecie sono stati utilizzati i dati provenienti dall’HVAC, da un sensore di temperatura interna dell'edificio e dal fotovoltaico installato nella struttura. La comunicazione tra il sistema distribuito e l’entità dell’HVAC avviene mediante il canale di comunicazione ModBus, mentre per quanto riguarda i dati della temperatura interna e del fotovoltaico questi vengono collezionati da sensori che inviano le informazioni sfruttando un canale di comunicazione che utilizza il protocollo MQTT, e lo stesso viene utilizzato come principale metodo di comunicazione all’interno del sistema, appoggiandosi ad un broker di messaggistica con modello publish/subscribe. L'automatizzazione del sistema è dovuta anche all'utilizzo di un modello di predizione con lo scopo di predire in maniera quanto più accurata possibile la temperatura interna all'edificio delle ore future. Per quanto riguarda il modello di predizione da me implementato e integrato nel sistema la scelta è stata quella di ispirarmi ad un modello ideato da Google nel 2014 ovvero il Sequence to Sequence. Il modello sviluppato si struttura come un encoder-decoder che utilizza le RNN, in particolare le reti LSTM.

Flower cover and bee diversity estimate by drone RGB images

Worldwide, biodiversity is decreasing due to climate change, habitat fragmentation and agricultural intensification. Bees are essential crops pollinator, but their abundance and diversity are decreasing as well. For their conservation, it is necessary to assess the status of bee population. Field data collection methods are expensive and time consuming thus, recently, new methods based on remote sensing are used. In this study we tested the possibility of using flower cover diversity estimated by UAV images (FCD-UAV) to assess bee diversity and abundance in 10 agricultural meadows in the Netherlands. In order to do so, field data of flower and bee diversity and abundance were collected during a campaign in May 2021. Furthermore, RGB images of the areas have been collected using Unmanned Aerial Vehicle (UAV) and post-processed into orthomosaics. Lastly, Random Forest machine learning algorithm was applied to estimate FCD of the species detected in each field. Resulting FCD was expressed with Shannon and Simpson diversity indices, which were successively correlated to bee Shannon and Simpson diversity indices, abundance and species richness. The results showed a positive relationship between FCD-UAV and in-situ collected data about bee diversity, evaluated with Shannon index, abundance and species richness. The strongest relationship was found between FCD (Shannon Index) and bee abundance with R2=0.52. Following, good correlations were found with bee species richness (R2=0.39) and bee diversity (R2=0.37). R2 values of the relationship between FCD (Simpson Index) and bee abundance, species richness and diversity were slightly inferior (0.45, 0.37 and 0.35, respectively). Our results suggest that the proposed method based on the coupling of UAV imagery and machine learning for the assessment of flower species diversity could be developed into valuable tools for large-scale, standardized and cost-effective monitoring of flower cover and of the habitat quality for bees.

Iis--integrated Interactome System: A Web-based Platform For The Annotation, Analysis And Visualization Of Protein-metabolite-gene-drug Interactions By Integrating A Variety Of Data Sources And Tools.

High-throughput screening of physical, genetic and chemical-genetic interactions brings important perspectives in the Systems Biology field, as the analysis of these interactions provides new insights into protein/gene function, cellular metabolic variations and the validation of therapeutic targets and drug design. However, such analysis depends on a pipeline connecting different tools that can automatically integrate data from diverse sources and result in a more comprehensive dataset that can be properly interpreted. We describe here the Integrated Interactome System (IIS), an integrative platform with a web-based interface for the annotation, analysis and visualization of the interaction profiles of proteins/genes, metabolites and drugs of interest. IIS works in four connected modules: (i) Submission module, which receives raw data derived from Sanger sequencing (e.g. two-hybrid system); (ii) Search module, which enables the user to search for the processed reads to be assembled into contigs/singlets, or for lists of proteins/genes, metabolites and drugs of interest, and add them to the project; (iii) Annotation module, which assigns annotations from several databases for the contigs/singlets or lists of proteins/genes, generating tables with automatic annotation that can be manually curated; and (iv) Interactome module, which maps the contigs/singlets or the uploaded lists to entries in our integrated database, building networks that gather novel identified interactions, protein and metabolite expression/concentration levels, subcellular localization and computed topological metrics, GO biological processes and KEGG pathways enrichment. This module generates a XGMML file that can be imported into Cytoscape or be visualized directly on the web. We have developed IIS by the integration of diverse databases following the need of appropriate tools for a systematic analysis of physical, genetic and chemical-genetic interactions. IIS was validated with yeast two-hybrid, proteomics and metabolomics datasets, but it is also extendable to other datasets. IIS is freely available online at: http://www.lge.ibi.unicamp.br/lnbio/IIS/.