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.

Methods for monitoring Machine Learning models

The following thesis aims to investigate the issues concerning the maintenance of a Machine Learning model over time, both about the versioning of the model itself and the data on which it is trained and about data monitoring tools and their distribution. The themes of Data Drift and Concept Drift were then explored and the performance of some of the most popular techniques in the field of Anomaly detection, such as VAE, PCA, and Monte Carlo Dropout, were evaluated.

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.

Analysis and improvements of an existing industrial hot stamping machine

The importance of product presentation in the marketing industry is well known. Labels are crucial for providing information to the buyer, but at a modest additional expense, a beautiful label with exquisite embellishments may also give the goods a sensation of high quality and elegance. Enhancing the capabilities of stamping machines is required to keep up with the increasing velocity of the production lines in the modern manufacturing industry and to offer new opportunities for customization. It’s in this context of improvements and refinements that this work takes place. The thesis was developed during an internship at Studio D, the firm that designs the mechanics of the machines produced by Cartes. The The aim of this work is to study possible upgrades for the existing hot stamping machines. The main focus of this work is centred on two objectives: first, evaluating the pressing forces generated by this machine and characterising how the mat used in the stamping process reacts to such forces. Second, propose a new conformation for the press mechanism in order to improve the rigidity and performance of the machines. The first objective is reached through a combined approach: the mat is crudely characterized with experimental data, while the frame of the machine is studied through FEM analysis. The results obtained are combined and used to upgrade a worksheet that allows to estimate the forces exerted by the machines. The second objective is reached with the proposal of new, improved designs for the main components of the machines.

Machine Learning algorithm on seat comfort level measurement, mapping pressure values with body regions

The comfort level of the seat has a major effect on the usage of a vehicle; thus, car manufacturers have been working on elevating car seat comfort as much as possible. However, still, the testing and evaluation of comfort are done using exhaustive trial and error testing and evaluation of data. In this thesis, we resort to machine learning and Artificial Neural Networks (ANN) to develop a fully automated approach. Even though this approach has its advantages in minimizing time and using a large set of data, it takes away the degree of freedom of the engineer on making decisions. The focus of this study is on filling the gap in a two-step comfort level evaluation which used pressure mapping with body regions to evaluate the average pressure supported by specific body parts and the Self-Assessment Exam (SAE) questions on evaluation of the person’s interest. This study has created a machine learning algorithm that works on giving a degree of freedom to the engineer in making a decision when mapping pressure values with body regions using ANN. The mapping is done with 92% accuracy and with the help of a Graphical User Interface (GUI) that facilitates the process during the testing time of comfort level evaluation of the car seat, which decreases the duration of the test analysis from days to hours.

Neural Machine Translation in the Field of Law. Analysis and Evaluation against Human Reference of NMT Raw Outputs of an Atto Costitutivo of S.r.l. from Italian into German.

Artificial Intelligence (AI) is gaining ever more ground in every sphere of human life, to the point that it is now even used to pass sentences in courts. The use of AI in the field of Law is however deemed quite controversial, as it could provide more objectivity yet entail an abuse of power as well, given that bias in algorithms behind AI may cause lack of accuracy. As a product of AI, machine translation is being increasingly used in the field of Law too in order to translate laws, judgements, contracts, etc. between different languages and different legal systems. In the legal setting of Company Law, accuracy of the content and suitability of terminology play a crucial role within a translation task, as any addition or omission of content or mistranslation of terms could entail legal consequences for companies. The purpose of the present study is to first assess which neural machine translation system between DeepL and ModernMT produces a more suitable translation from Italian into German of the atto costitutivo of an Italian s.r.l. in terms of accuracy of the content and correctness of terminology, and then to assess which translation proves to be closer to a human reference translation. In order to achieve the above-mentioned aims, two human and automatic evaluations are carried out based on the MQM taxonomy and the BLEU metric. Results of both evaluations show an overall better performance delivered by ModernMT in terms of content accuracy, suitability of terminology, and closeness to a human translation. As emerged from the MQM-based evaluation, its accuracy and terminology errors account for just 8.43% (as opposed to DeepL’s 9.22%), while it obtains an overall BLEU score of 29.14 (against DeepL’s 27.02). The overall performances however show that machines still face barriers in overcoming semantic complexity, tackling polysemy, and choosing domain-specific terminology, which suggests that the discrepancy with human translation may still be remarkable.

Ma il machine interpreting sogna pecore elettroniche? Case study del dispositivo WT2 Plus in contesto reale

Il presente elaborato analizza le rese del sistema di interpretazione automatica WT2 Plus di Timekettle in un reale contesto didattico. Nello specifico, sono state condotte tre sperimentazioni presso l’Accademia Europea di Manga utilizzando due modalità di interpretazione supportate dal dispositivo. L’obiettivo è valutare la qualità di un sistema di interpretazione automatica in un contesto reale, dato che allo stato attuale mancano studi che valutino la performance di questi dispositivi nei reali contesti per i quali sono stati sviluppati. Il primo capitolo ripercorre la storia dell’interpretazione automatica, cui fa seguito la spiegazione della tecnologia alla base dei sistemi disponibili sul mercato, di cui poi si presenta lo stato dell’arte. Successivamente si delinea in breve il prossimo passaggio evolutivo nell’interpretazione automatica e si analizzano tre casi studio simili a quello proposto. Il capitolo si conclude con l’approccio scelto per valutare la performance del dispositivo. Il secondo capitolo, dedicato alla metodologia, si apre con una panoramica su WT2 Plus e sull’azienda produttrice per poi descrivere il metodo usato nelle sperimentazioni. Nello specifico, le sperimentazioni sono state condotte durante alcune delle attività previste nel normale svolgimento delle lezioni presso l’Accademia Europea di Manga con il coinvolgimento di uno studente anglofono e di due membri del personale accademico. I dati raccolti sono analizzati e discussi rispettivamente nei capitoli 3 e 4. I risultati ottenuti sembrano suggerire che il dispositivo non sia, allo stato attuale, compatibile con un reale contesto didattico, per via del quasi mancato funzionamento di una delle modalità prescelte, per il servizio di interpretazione, che risulta ancora troppo letterale, la monotonia dell’intonazione della voce automatica e infine per la traduzione quasi completamente errata della terminologia tecnica.

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.

Asynchronous Massive Machine-Type Access with Massive MIMO

In this thesis asynchronous contention resolution diversity slotted ALOHA (ACRDA) is studied and implemented on computer to simulate a typical massive IoT scenario. Chapter 1 gives a general overview of existing multiple access schemes, reporting their fundamental concepts focusing more on Coded Random Access schemes and their characteristics. In Chapter 2 the asynchronous protocol ACRDA is explained in depth analyzing all parts of the scheme. In the third Chapter the results obtained following various simulations of the asynchronous scheme are reported and their performance are analyzed.

Dentin Bond Strength Of A Fluoride-releasing Adhesive System Submitted To Ph-cycling.

To evaluate the microtensile bond strength (µTBS) of a fluoride-containing adhesive system submitted to a pH-cycling and storage time regimen for primary outcomes. As secondary outcomes the fluoride released amount was evaluated. Twelve dentin surfaces from sound third molar were divided into 2 groups according to adhesive systems: Clearfil SE Protect (PB) and Clearfil SE Bond (SE). Sticks obtained (1.0 mm2) from teeth were randomly divided into 3 subgroups according to storage regimen model: immediate (24h); 5-month deionized water (W); and pH-cycling model (C). All sticks were tested for µTBS in a universal testing machine. Fluoride concentration was obtained from 1-4 days and 30-day in W and 1-4 days in demineralization (DE)/remineralization (RE) solutions from C, using a fluoride-specific electrode. µTBS and fluoride released data were, respectively, submitted to ANOVA in a split plot design and Tukey, and Friedman' tests (a=0.05). There was no significant interaction between adhesive system and storage regimen for µTBS. W showed the lowest µTBS values. There was no significant difference between 24 h and C models for µTBS. There was no significant difference between adhesive systems. Failure mode was predominantly cohesive within composite for the 24 h and W, for the C group it was mixed for SE and cohesive within composite for PB adhesive system. Fluoride concentrations in the DE/RE solutions were less than 0.03125 ppm and not detected in W. In conclusion, the fluoride-containing adhesive system performed similarly to the regular one. Hydrolytic degradation is the main problem with both adhesive systems, regardless of fluoride contents.

In vitro evaluation of an alternative method to bond molar tubes

Despite the advances in bonding materials, many clinicians today still prefer to place bands on molar teeth. Molar bonding procedures need improvement to be widely accepted clinically. OBJECTIVE: The purpose of this study was to evaluate the shear bond strength when an additional adhesive layer was applied on the occlusal tooth/tube interface to provide reinforcement to molar tubes. MATERIAL AND METHODS: Sixty third molars were selected and allocated to the 3 groups: group 1 received a conventional direct bond followed by the application of an additional layer of adhesive on the occlusal tooth/tube interface, group 2 received a conventional direct bond, and group 3 received a conventional direct bond and an additional cure time of 10 s. The specimens were debonded in a universal testing machine. The results were analyzed statistically by ANOVA and Tukey's test (α=0.05). RESULTS: Group 1 had a significantly higher (p<0.05) shear bond strength compared to groups 2 and 3. No difference was detected between groups 2 and 3 (p>0.05). CONCLUSIONS: The present in vitro findings indicate that the application of an additional layer of adhesive on the tooth/tube interface increased the shear bond strength of the bonded molar tubes.

Bonding agent underneath sealant: shear bond strength to oil-contaminated

This study evaluated in vitro the shear bond strength of a resin-based pit-and-fissure sealant (Fluroshield - F) associated with either an ethanol-based (Adper Single Bond 2 - SB) or an acetone-based (Prime & Bond - PB) adhesive system under conditions of oil contamination. Mesial and distal enamel surfaces from 30 sound third molars were randomly assigned to 2 groups (n=30): I - no oil contamination; II - oil contamination. Contamination (0.25 mL during 10 s) was performed after 37% phosphoric acid etching with an air/oil spray. The specimens were randomly assigned to subgroups, according to the bonding protocol adopted: subgroup A - F was applied to enamel without an intermediate bonding agent layer; In subgroups B and C, SB and PB, respectively, were applied, light-cured, and then F was applied and light-cured. Shear bond strength was tested at a crosshead speed of 0.5 mm/min in a universal testing machine. Means (± SD) in MPa were: IA-11.28 (±1.84); IIA-12.02 (±1.15); IB-9.73 (±2.38); IIB-9.62 (±2.29); IC-28.30 (±1.63); and IIC-25.50 (±1.91). It may be concluded that the oil contamination affected negatively the sealant bonding to enamel and the acetone-based adhesive system (PB) layer applied underneath the sealant was able to prevent its deleterious effects to adhesion.

Effect of casting atmosphere on the shear bond strength of a ceramic to Ni-Cr and Ni-Cr-Be alloys

The success of metal-ceramic restorations depends on an optimal bond between metal and ceramic. This study evaluated the effect of 3 casting atmospheres on the metal-ceramic bond strength (MCBS) of 2 Ni-Cr alloys, with beryllium (Fit Cast V) and without beryllium (Fit Cast SB). Sixty acrylic resin patterns (8 mm long and 5 mm diameter) were obtained using a fluorocarbon resin matrix. Wax was used to refine the surface of acrylic resin patterns that were invested and cast in an induction casting machine under normal, vacuum, and argon atmospheres at a temperature of 1340ºC. The castings were divested manually and airborne-particle abraded with 100-µm aluminum-oxide. Ten castings were obtained for each group. The IPS Classic V ceramic was applied (2 mm high and 5 mm diameter). The shear bond strength was tested in a mechanical testing machine with a crosshead speed of 2.0 mm/min. The MCBS data (MPa) were subjected to 2-way analysis of variance (α=0.05). There was no statistically significant difference (p>0.05) between the alloys or among the casting atmospheres. Within the limitations of this study, it may be concluded that the presence of beryllium and the casting atmosphere did not interfere in the MCBS of the evaluated metal-ceramic combinations

Comparative study of chemical and mechanical retentive systems for bonding of indirect composite resin to commercially pure titanium

This study evaluated the effect of chemical and mechanical surface treatments for cast metal alloys on the bond strength of an indirect composite resin (Artglass) to commercially pure titanium (cpTi). Thirty cylindrical metal rods (3 mm diameter x 60 mm long) were cast in grade-1 cpTi and randomly assigned to 6 groups (n=5) according to the received surface treatment: sandblasting; chemical treatment; mechanical treatment - 0.4 mm beads; mechanical treatment - 0.6 mm beads; chemical/mechanical treatment - 0.4 mm; and chemical/mechanical treatment - 0.6 mm beads. Artglass rings (6.0 mm diameter x 2.0 mm thick) were light cured around the cpTi rods, according manufacturer's specifications. The specimens were invested in hard gypsum and their bond strength (in MPa) to the rods was measured at fracture with a universal testing machine at a crosshead speed of 2.0 mm/min and 500 kgf load cell. Data were analyzed statistically by one-way ANOVA and Tukey test (a=5%). The surface treatments differed significantly from each other (p<0.05) regarding the recorded bond strengths. Chemical retention and sandblasting showed statistically similar results to each other (p=0.139) and both had significantly lower bond strengths (p<0.05) than the other treatments. In conclusion, mechanical retention, either associated or not to chemical treatment, provided higher bond strength of the indirect composite resin to cpTi.

Bond strength of a pit-and-fissure sealant associated to etch-and-rinse and self-etching adhesive systems to saliva-contaminated enamel: individual vs. simultaneous light curing

This study evaluated in vitro the shear bond strength (SBS) of a resin-based pit-and-fissure sealant [Fluroshield (F), Dentsply/Caulk] associated with either an etch-and-rinse [Adper Single Bond 2 (SB), 3M/ESPE] or a self-etching adhesive system [Clearfil S3 Bond (S3), Kuraray Co., Ltd.] to saliva-contaminated enamel, comparing two curing protocols: individual light curing of the adhesive system and the sealant or simultaneous curing of both materials. Mesial and distal enamel surfaces from 45 sound third molars were randomly assigned to 6 groups (n=15), according to the bonding technique: I - F was applied to 37% phosphoric acid etched enamel. The other groups were contaminated with fresh human saliva (0.01 mL; 10 s) after acid etching: II - SB and F were light cured separately; III - SB and F were light cured together; IV - S3 and F were light cured separately; V - S3 and F were light cured simultaneously; VI - F was applied to saliva-contaminated, acid-etched enamel without an intermediate bonding agent layer. SBS was tested to failure in a universal testing machine at 0.5 mm/min. Data were analyzed by one-way ANOVA and Fisher's test (α=0.05).The debonded specimens were examined with a stereomicroscope to assess the failure modes. Three representative specimens from each group were observed under scanning electron microscopy for a qualitative analysis. Mean SBS in MPa were: I-12.28 (±4.29); II-8.57 (±3.19); III-7.97 (±2.16); IV-12.56 (±3.11); V-11.45 (±3.77); and VI-7.47 (±1.99). In conclusion, individual or simultaneous curing of the intermediate bonding agent layer and the resin sealant did not seem to affect bond strength to saliva-contaminated enamel. S3/F presented significantly higher SBS than the that of the groups treated with SB etch-and-rinse adhesive system and similar SBS to that of the control group, in which the sealant was applied under ideal dry, noncontaminated conditions.