Machine learning as a service for high energy physics (MLaaS4HEP): a service for ML-based data analyses

With the CERN LHC program underway, there has been an acceleration of data growth in the High Energy Physics (HEP) field and the usage of Machine Learning (ML) in HEP will be critical during the HL-LHC program when the data that will be produced will reach the exascale. ML techniques have been successfully used in many areas of HEP nevertheless, the development of a ML project and its implementation for production use is a highly time-consuming task and requires specific skills. Complicating this scenario is the fact that HEP data is stored in ROOT data format, which is mostly unknown outside of the HEP community. The work presented in this thesis is focused on the development of a ML as a Service (MLaaS) solution for HEP, aiming to provide a cloud service that allows HEP users to run ML pipelines via HTTP calls. These pipelines are executed by using the MLaaS4HEP framework, which allows reading data, processing data, and training ML models directly using ROOT files of arbitrary size from local or distributed data sources. Such a solution provides HEP users non-expert in ML with a tool that allows them to apply ML techniques in their analyses in a streamlined manner. Over the years the MLaaS4HEP framework has been developed, validated, and tested and new features have been added. A first MLaaS solution has been developed by automatizing the deployment of a platform equipped with the MLaaS4HEP framework. Then, a service with APIs has been developed, so that a user after being authenticated and authorized can submit MLaaS4HEP workflows producing trained ML models ready for the inference phase. A working prototype of this service is currently running on a virtual machine of INFN-Cloud and is compliant to be added to the INFN Cloud portfolio of services.

Considerations on indirect adhesive restorations at the material and tooth interfaces

Objective: Lithium-silicate (LiSi) ceramic is nowadays widely used in dentistry. However, for the longevity of LiSi indirect restorations, it is important to pretreat the material and the dental substrate adequately. However, is not certain how the simplification of the manufacturing and conditioning procedures influences the bonding performances of LiSi ceramic restorations. Accordingly, the aims of this thesis were to investigate the effect of: 1) different LiSi ceramic surface decontamination procedures on the shear bond strength (SBS) to resin composite; 2) different types of lithium-disilicate (LiDi) (pressed vs CAD-CAM) on SBS to resin composite; 3) an experimental metal salt-based zirconium oxynitrate etchant [ZrO(NO3)2] on bonding performances to dentin. Materials and Methods: SBS test was used to investigate the influence of different cleaning protocols applied, or different processing techniques (CAD or PRESS) on the bond strength to composite resin. The third study tackled the interface between restorative materials and dentin, and investigated the microtensile bond strength test (µTBS), nanoleakage expression analysis (NL), gelatin zymography and in situ zymography of dentin conditioned with an experimental metal salt-based zirconium oxynitrate etchant [ZrO(NO3)2]. Results: MEP showed comparable bond strength to the double HP etching and higher compared to other groups. BS of press LiSi to composite was higher than that of CAD/CAM LiSi. ZON pretreatment increased bond strength to dentin when used with a universal adhesive, and inhibited dentinal endogenous enzymes. Conclusions: While simplification of the LiSi conditioning and cleaning procedures seems to yield bond strength comparable to the traditional procedures, it could be recommended in the clinical practice. However, pressed LiSi still seems to perform better in terms of bond strength compared to the CAD/CAM LiSi. Further, the novel ZON etchant seems to perform better compared to the traditional phosphoric dentin etching.

Uncovering the encoding and transmission of behaviourally relevant information in neural activity

Most cognitive functions require the encoding and routing of information across distributed networks of brain regions. Information propagation is typically attributed to physical connections existing between brain regions, and contributes to the formation of spatially correlated activity patterns, known as functional connectivity. While structural connectivity provides the anatomical foundation for neural interactions, the exact manner in which it shapes functional connectivity is complex and not yet fully understood. Additionally, traditional measures of directed functional connectivity only capture the overall correlation between neural activity, and provide no insight on the content of transmitted information, limiting their ability in understanding neural computations underlying the distributed processing of behaviorally-relevant variables. In this work, we first study the relationship between structural and functional connectivity in simulated recurrent spiking neural networks with spike timing dependent plasticity. We use established measures of time-lagged correlation and overall information propagation to infer the temporal evolution of synaptic weights, showing that measures of dynamic functional connectivity can be used to reliably reconstruct the evolution of structural properties of the network. Then, we extend current methods of directed causal communication between brain areas, by deriving an information-theoretic measure of Feature-specific Information Transfer (FIT) quantifying the amount, content and direction of information flow. We test FIT on simulated data, showing its key properties and advantages over traditional measures of overall propagated information. We show applications of FIT to several neural datasets obtained with different recording methods (magneto and electro-encephalography, spiking activity, local field potentials) during various cognitive functions, ranging from sensory perception to decision making and motor learning. Overall, these analyses demonstrate the ability of FIT to advance the investigation of communication between brain regions, uncovering the previously unaddressed content of directed information flow.

Quality of Service in Distributed Stream Processing for large scale Smart Pervasive Environments

The wide diffusion of cheap, small, and portable sensors integrated in an unprecedented large variety of devices and the availability of almost ubiquitous Internet connectivity make it possible to collect an unprecedented amount of real time information about the environment we live in. These data streams, if properly and timely analyzed, can be exploited to build new intelligent and pervasive services that have the potential of improving people's quality of life in a variety of cross concerning domains such as entertainment, health-care, or energy management. The large heterogeneity of application domains, however, calls for a middleware-level infrastructure that can effectively support their different quality requirements. In this thesis we study the challenges related to the provisioning of differentiated quality-of-service (QoS) during the processing of data streams produced in pervasive environments. We analyze the trade-offs between guaranteed quality, cost, and scalability in streams distribution and processing by surveying existing state-of-the-art solutions and identifying and exploring their weaknesses. We propose an original model for QoS-centric distributed stream processing in data centers and we present Quasit, its prototype implementation offering a scalable and extensible platform that can be used by researchers to implement and validate novel QoS-enforcement mechanisms. To support our study, we also explore an original class of weaker quality guarantees that can reduce costs when application semantics do not require strict quality enforcement. We validate the effectiveness of this idea in a practical use-case scenario that investigates partial fault-tolerance policies in stream processing by performing a large experimental study on the prototype of our novel LAAR dynamic replication technique. Our modeling, prototyping, and experimental work demonstrates that, by providing data distribution and processing middleware with application-level knowledge of the different quality requirements associated to different pervasive data flows, it is possible to improve system scalability while reducing costs.

Parallel modeling of the electric field distribution in the brain

The term "Brain Imaging" identi�es a set of techniques to analyze the structure and/or functional behavior of the brain in normal and/or pathological situations. These techniques are largely used in the study of brain activity. In addition to clinical usage, analysis of brain activity is gaining popularity in others recent �fields, i.e. Brain Computer Interfaces (BCI) and the study of cognitive processes. In this context, usage of classical solutions (e.g. f MRI, PET-CT) could be unfeasible, due to their low temporal resolution, high cost and limited portability. For these reasons alternative low cost techniques are object of research, typically based on simple recording hardware and on intensive data elaboration process. Typical examples are ElectroEncephaloGraphy (EEG) and Electrical Impedance Tomography (EIT), where electric potential at the patient's scalp is recorded by high impedance electrodes. In EEG potentials are directly generated from neuronal activity, while in EIT by the injection of small currents at the scalp. To retrieve meaningful insights on brain activity from measurements, EIT and EEG relies on detailed knowledge of the underlying electrical properties of the body. This is obtained from numerical models of the electric �field distribution therein. The inhomogeneous and anisotropic electric properties of human tissues make accurate modeling and simulation very challenging, leading to a tradeo�ff between physical accuracy and technical feasibility, which currently severely limits the capabilities of these techniques. Moreover elaboration of data recorded requires usage of regularization techniques computationally intensive, which influences the application with heavy temporal constraints (such as BCI). This work focuses on the parallel implementation of a work-flow for EEG and EIT data processing. The resulting software is accelerated using multi-core GPUs, in order to provide solution in reasonable times and address requirements of real-time BCI systems, without over-simplifying the complexity and accuracy of the head models.

Advanced Beamforming for Distributed and Multi-Beam Networks

Beamforming entails joint processing of multiple signals received or transmitted by an array of antennas. This thesis addresses the implementation of beamforming in two distinct systems, namely a distributed network of independent sensors, and a broad-band multi-beam satellite network. With the rising popularity of wireless sensors, scientists are taking advantage of the flexibility of these devices, which come with very low implementation costs. Simplicity, however, is intertwined with scarce power resources, which must be carefully rationed to ensure successful measurement campaigns throughout the whole duration of the application. In this scenario, distributed beamforming is a cooperative communication technique, which allows nodes in the network to emulate a virtual antenna array seeking power gains in the order of the size of the network itself, when required to deliver a common message signal to the receiver. To achieve a desired beamforming configuration, however, all nodes in the network must agree upon the same phase reference, which is challenging in a distributed set-up where all devices are independent. The first part of this thesis presents new algorithms for phase alignment, which prove to be more energy efficient than existing solutions. With the ever-growing demand for broad-band connectivity, satellite systems have the great potential to guarantee service where terrestrial systems can not penetrate. In order to satisfy the constantly increasing demand for throughput, satellites are equipped with multi-fed reflector antennas to resolve spatially separated signals. However, incrementing the number of feeds on the payload corresponds to burdening the link between the satellite and the gateway with an extensive amount of signaling, and to possibly calling for much more expensive multiple-gateway infrastructures. This thesis focuses on an on-board non-adaptive signal processing scheme denoted as Coarse Beamforming, whose objective is to reduce the communication load on the link between the ground station and space segment.

Ultra-low power WSNs: distributed signal processing and dynamic resource management

This thesis presents several data processing and compression techniques capable of addressing the strict requirements of wireless sensor networks. After introducing a general overview of sensor networks, the energy problem is introduced, dividing the different energy reduction approaches according to the different subsystem they try to optimize. To manage the complexity brought by these techniques, a quick overview of the most common middlewares for WSNs is given, describing in detail SPINE2, a framework for data processing in the node environment. The focus is then shifted on the in-network aggregation techniques, used to reduce data sent by the network nodes trying to prolong the network lifetime as long as possible. Among the several techniques, the most promising approach is the Compressive Sensing (CS). To investigate this technique, a practical implementation of the algorithm is compared against a simpler aggregation scheme, deriving a mixed algorithm able to successfully reduce the power consumption. The analysis moves from compression implemented on single nodes to CS for signal ensembles, trying to exploit the correlations among sensors and nodes to improve compression and reconstruction quality. The two main techniques for signal ensembles, Distributed CS (DCS) and Kronecker CS (KCS), are introduced and compared against a common set of data gathered by real deployments. The best trade-off between reconstruction quality and power consumption is then investigated. The usage of CS is also addressed when the signal of interest is sampled at a Sub-Nyquist rate, evaluating the reconstruction performance. Finally the group sparsity CS (GS-CS) is compared to another well-known technique for reconstruction of signals from an highly sub-sampled version. These two frameworks are compared again against a real data-set and an insightful analysis of the trade-off between reconstruction quality and lifetime is given.

Residual visual processing following real or virtual lesions to primary visual pathways

Lesions to the primary geniculo-striate visual pathway cause blindness in the contralesional visual field. Nevertheless, previous studies have suggested that patients with visual field defects may still be able to implicitly process the affective valence of unseen emotional stimuli (affective blindsight) through alternative visual pathways bypassing the striate cortex. These alternative pathways may also allow exploitation of multisensory (audio-visual) integration mechanisms, such that auditory stimulation can enhance visual detection of stimuli which would otherwise be undetected when presented alone (crossmodal blindsight). The present dissertation investigated implicit emotional processing and multisensory integration when conscious visual processing is prevented by real or virtual lesions to the geniculo-striate pathway, in order to further clarify both the nature of these residual processes and the functional aspects of the underlying neural pathways. The present experimental evidence demonstrates that alternative subcortical visual pathways allow implicit processing of the emotional content of facial expressions in the absence of cortical processing. However, this residual ability is limited to fearful expressions. This finding suggests the existence of a subcortical system specialised in detecting danger signals based on coarse visual cues, therefore allowing the early recruitment of flight-or-fight behavioural responses even before conscious and detailed recognition of potential threats can take place. Moreover, the present dissertation extends the knowledge about crossmodal blindsight phenomena by showing that, unlike with visual detection, sound cannot crossmodally enhance visual orientation discrimination in the absence of functional striate cortex. This finding demonstrates, on the one hand, that the striate cortex plays a causative role in crossmodally enhancing visual orientation sensitivity and, on the other hand, that subcortical visual pathways bypassing the striate cortex, despite affording audio-visual integration processes leading to the improvement of simple visual abilities such as detection, cannot mediate multisensory enhancement of more complex visual functions, such as orientation discrimination.

Computational Investigation of the Molecular Mechanisms Regulating Nucleic Acid Processing Metalloenzymes

Polymerases and nucleases are enzymes processing DNA and RNA. They are involved in crucial processes for cell life by performing the extension and the cleavage of nucleic acid chains during genome replication and maintenance. Additionally, both enzymes are often associated to several diseases, including cancer. In order to catalyze the reaction, most of them operate via the two-metal-ion mechanism. For this, despite showing relevant differences in structure, function and catalytic properties, they share common catalytic elements, which comprise the two catalytic ions and their first-shell acidic residues. Notably, recent studies of different metalloenzymes revealed the recurrent presence of additional elements surrounding the active site, thus suggesting an extended two-metal-ion-centered architecture. However, whether these elements have a catalytic function and what is their role is still unclear. In this work, using state-of-the-art computational techniques, second- and third-shell elements are showed to act in metallonucleases favoring the substrate positioning and leaving group release. In particular, in hExo1 a transient third metal ion is recruited and positioned near the two-metal-ion site by a structurally conserved acidic residue to assist the leaving group departure. Similarly, in hFEN1 second- and third-shell Arg/Lys residues operate the phosphate steering mechanism through (i) substrate recruitment, (ii) precise cleavage localization, and (iii) leaving group release. Importantly, structural comparisons of hExo1, hFEN1 and other metallonucleases suggest that similar catalytic mechanisms may be shared by other enzymes. Overall, the results obtained provide an extended vision on parallel strategies adopted by metalloenzymes, which employ divalent metal ion or positively charged residues to ensure efficient and specific catalysis. Furthermore, these outcomes may have implications for de novo enzyme engineering and/or drug design to modulate nucleic acid processing.