Education in Computer Science and the Role of the Teacher in the Environment of Open Educational Resources

Report published in the Proceedings of the National Conference on "Education in the Information Society", Plovdiv, May, 2013

The applied computer science of economic analyses: Internet prioritization and Internet pricing, and software engineering and software pricing for the mass market

This research examines evolving issues in applied computer science and applies economic and business analyses as well. There are two main areas. The first is internetwork communications as embodied by the Internet. The goal of the research is to devise an efficient pricing, prioritization, and incentivization plan that could be realistically implemented on the existing infrastructure. Criteria include practical and economic efficiency, and proper incentives for both users and providers. Background information on the evolution and functional operation of the Internet is given, and relevant literature is surveyed and analyzed. Economic analysis is performed on the incentive implications of the current pricing structure and organization. The problems are identified, and minimally disruptive solutions are proposed for all levels of implementation to the lowest level protocol. Practical issues are considered and performance analyses are done. The second area of research is mass market software engineering, and how this differs from classical software engineering. Software life-cycle revenues are analyzed and software pricing and timing implications are derived. A profit maximizing methodology is developed to select or defer the development of software features for inclusion in a given release. An iterative model of the stages of the software development process is developed, taking into account new communications capabilities as well as profitability. ^

An innovative approach to teaching structural induction for computer science

Proofs by induction are central to many computer science areas such as data structures, theory of computation, programming languages, program efficiency-time complexity, and program correctness. Proofs by induction can also improve students’ understanding and performance of computer science concepts such as programming languages, algorithm design, and recursion, as well as serve as a medium for teaching them. Even though students are exposed to proofs by induction in many courses of their curricula, they still have difficulties understanding and performing them. This impacts the whole course of their studies, since proofs by induction are omnipresent in computer science. Specifically, students do not gain conceptual understanding of induction early in the curriculum and as a result, they have difficulties applying it to more advanced areas later on in their studies. The goal of my dissertation is twofold: (1) identifying sources of computer science students’ difficulties with proofs by induction, and (2) developing a new approach to teaching proofs by induction by way of an interactive and multimodal electronic book (e-book). For the first goal, I undertook a study to identify possible sources of computer science students’ difficulties with proofs by induction. Its results suggest that there is a close correlation between students’ understanding of inductive definitions and their understanding and performance of proofs by induction. For designing and developing my e-book, I took into consideration the results of my study, as well as the drawbacks of the current methodologies of teaching proofs by induction for computer science. I designed my e-book to be used as a standalone and complete educational environment. I also conducted a study on the effectiveness of my e-book in the classroom. The results of my study suggest that, unlike the current methodologies of teaching proofs by induction for computer science, my e-book helped students overcome many of their difficulties and gain conceptual understanding of proofs induction.

An Innovative Approach to Teaching Structural Induction for Computer Science

Proofs by induction are central to many computer science areas such as data structures, theory of computation, programming languages, program efficiency-time complexity, and program correctness. Proofs by induction can also improve students’ understanding of and performance with computer science concepts such as programming languages, algorithm design, and recursion, as well as serve as a medium for teaching them. Even though students are exposed to proofs by induction in many courses of their curricula, they still have difficulties understanding and performing them. This impacts the whole course of their studies, since proofs by induction are omnipresent in computer science. Specifically, students do not gain conceptual understanding of induction early in the curriculum and as a result, they have difficulties applying it to more advanced areas later on in their studies. The goal of my dissertation is twofold: 1. identifying sources of computer science students’ difficulties with proofs by induction, and 2. developing a new approach to teaching proofs by induction by way of an interactive and multimodal electronic book (e-book). For the first goal, I undertook a study to identify possible sources of computer science students’ difficulties with proofs by induction. Its results suggest that there is a close correlation between students’ understanding of inductive definitions and their understanding and performance of proofs by induction. For designing and developing my e-book, I took into consideration the results of my study, as well as the drawbacks of the current methodologies of teaching proofs by induction for computer science. I designed my e-book to be used as a standalone and complete educational environment. I also conducted a study on the effectiveness of my e-book in the classroom. The results of my study suggest that, unlike the current methodologies of teaching proofs by induction for computer science, my e-book helped students overcome many of their difficulties and gain conceptual understanding of proofs induction.

Periodic nonlinear Fourier transform for fiber-optic communications, Part I:theory and numerical methods

In this work, we introduce the periodic nonlinear Fourier transform (PNFT) method as an alternative and efficacious tool for compensation of the nonlinear transmission effects in optical fiber links. In the Part I, we introduce the algorithmic platform of the technique, describing in details the direct and inverse PNFT operations, also known as the inverse scattering transform for periodic (in time variable) nonlinear Schrödinger equation (NLSE). We pay a special attention to explaining the potential advantages of the PNFT-based processing over the previously studied nonlinear Fourier transform (NFT) based methods. Further, we elucidate the issue of the numerical PNFT computation: we compare the performance of four known numerical methods applicable for the calculation of nonlinear spectral data (the direct PNFT), in particular, taking the main spectrum (utilized further in Part II for the modulation and transmission) associated with some simple example waveforms as the quality indicator for each method. We show that the Ablowitz-Ladik discretization approach for the direct PNFT provides the best performance in terms of the accuracy and computational time consumption.

Solar coronal jets: observations, theory, and modeling

Coronal jets represent important manifestations of ubiquitous solar transients, which may be the source of significant mass and energy input to the upper solar atmosphere and the solar wind. While the energy involved in a jet-like event is smaller than that of “nominal” solar flares and coronal mass ejections (CMEs), jets share many common properties with these phenomena, in particular, the explosive magnetically driven dynamics. Studies of jets could, therefore, provide critical insight for understanding the larger, more complex drivers of the solar activity. On the other side of the size-spectrum, the study of jets could also supply important clues on the physics of transients close or at the limit of the current spatial resolution such as spicules. Furthermore, jet phenomena may hint to basic process for heating the corona and accelerating the solar wind; consequently their study gives us the opportunity to attack a broad range of solar-heliospheric problems.

Student attitudes to plagiarism and collusion within computer science

There is a widespread perception among staff in Computer Science that plagiarism is a major problem particularly in the form of collusion in programming exercises. While departments often make use of electronic detection measures, the time consumed prosecuting plagiarism offences remains a problem. As a result departments continue to seek ways to reduce the amount of plagiarism and collusion that occurs. This paper reports the findings of a questionnaire based study which attempted to assess the students' attitudes to the issues involved in the hope that such an understanding might result in practical measures for minimizing the problem. The study revealed that while students did understand the definition of plagiarism in its most extreme cases they were often confused about less clear-cut situations. Changes in the previous experience of incoming students meeting modules originally designed on the assumption that students already had some programming background and were equipped for self-directed study would also appear to be a contributory factor in the extent of collusion in programming exercises.

Non-Perturbative Methods in Quantum Field Theory and Quantum Gravity

This thesis considers non-perturbative methods in quantum field theory with applications to gravity and cosmology. In particular, there are chapters on black hole holography, inflationary model building, and the conformal bootstrap.

Radio Frequency Superconducting Quantum Interference Device Meta-atoms and Metamaterials: Experiment, Theory, and Analysis

Metamamterials are 1D, 2D or 3D arrays of articial atoms. The articial atoms, called "meta-atoms", can be any component with tailorable electromagnetic properties, such as resonators, LC circuits, nano particles, and so on. By designing the properties of individual meta-atoms and the interaction created by putting them in a lattice, one can create a metamaterial with intriguing properties not found in nature. My Ph. D. work examines the meta-atoms based on radio frequency superconducting quantum interference devices (rf-SQUIDs); their tunability with dc magnetic field, rf magnetic field, and temperature are studied. The rf-SQUIDs are superconducting split ring resonators in which the usual capacitance is supplemented with a Josephson junction, which introduces strong nonlinearity in the rf properties. At relatively low rf magnetic field, a magnetic field tunability of the resonant frequency of up to 80 THz/Gauss by dc magnetic field is observed, and a total frequency tunability of 100% is achieved. The macroscopic quantum superconducting metamaterial also shows manipulative self-induced broadband transparency due to a qualitatively novel nonlinear mechanism that is different from conventional electromagnetically induced transparency (EIT) or its classical analogs. A near complete disappearance of resonant absorption under a range of applied rf flux is observed experimentally and explained theoretically. The transparency comes from the intrinsic bi-stability and can be tuned on/ off easily by altering rf and dc magnetic fields, temperature and history. Hysteretic in situ 100% tunability of transparency paves the way for auto-cloaking metamaterials, intensity dependent filters, and fast-tunable power limiters. An rf-SQUID metamaterial is shown to have qualitatively the same behavior as a single rf-SQUID with regards to dc flux, rf flux and temperature tuning. The two-tone response of self-resonant rf-SQUID meta-atoms and metamaterials is then studied here via intermodulation (IM) measurement over a broad range of tone frequencies and tone powers. A sharp onset followed by a surprising strongly suppressed IM region near the resonance is observed. This behavior can be understood employing methods in nonlinear dynamics; the sharp onset, and the gap of IM, are due to sudden state jumps during a beat of the two-tone sum input signal. The theory predicts that the IM can be manipulated with tone power, center frequency, frequency difference between the two tones, and temperature. This quantitative understanding potentially allows for the design of rf-SQUID metamaterials with either very low or very high IM response.

Concepts and methods for efficient decentralized learning in federated settings

Deep Neural Networks (DNNs) have revolutionized a wide range of applications beyond traditional machine learning and artificial intelligence fields, e.g., computer vision, healthcare, natural language processing and others. At the same time, edge devices have become central in our society, generating an unprecedented amount of data which could be used to train data-hungry models such as DNNs. However, the potentially sensitive or confidential nature of gathered data poses privacy concerns when storing and processing them in centralized locations. To this purpose, decentralized learning decouples model training from the need of directly accessing raw data, by alternating on-device training and periodic communications. The ability of distilling knowledge from decentralized data, however, comes at the cost of facing more challenging learning settings, such as coping with heterogeneous hardware and network connectivity, statistical diversity of data, and ensuring verifiable privacy guarantees. This Thesis proposes an extensive overview of decentralized learning literature, including a novel taxonomy and a detailed description of the most relevant system-level contributions in the related literature for privacy, communication efficiency, data and system heterogeneity, and poisoning defense. Next, this Thesis presents the design of an original solution to tackle communication efficiency and system heterogeneity, and empirically evaluates it on federated settings. For communication efficiency, an original method, specifically designed for Convolutional Neural Networks, is also described and evaluated against the state-of-the-art. Furthermore, this Thesis provides an in-depth review of recently proposed methods to tackle the performance degradation introduced by data heterogeneity, followed by empirical evaluations on challenging data distributions, highlighting strengths and possible weaknesses of the considered solutions. Finally, this Thesis presents a novel perspective on the usage of Knowledge Distillation as a mean for optimizing decentralized learning systems in settings characterized by data heterogeneity or system heterogeneity. Our vision on relevant future research directions close the manuscript.

Upconversion in Nd(3+)-doped glasses: Microscopic theory and spectroscopic measurements

In this work, we report a systematic investigation of upconversion losses and their effects on fluorescence quantum efficiency and fractional thermal loading in Nd(3+)-doped fluoride glasses. The energy transfer upconversion (gamma(up)) parameter, which describes upconversion losses, was experimentally determined using different methods: thermal lens (TL) technique and steady state luminescence (SSL) measurements. Additionally, the upconversion parameter was also obtained from energy transfer models and excited state absorption measurements. The results reveal that the microscopic treatment provided by the energy transfer models is similar to the macroscopic ones achieved from the TL and SSL measurements because similar gamma(up) parameters were obtained. Besides, the achieved results also point out the migration-assisted energy transfer according to diffusion-limited regime rather than hopping regime as responsible for the upconversion losses in Nd-doped glasses. (c) 2008 American Institute of Physics.

Density functional theory and ab initio molecular dynamics study of NO adsorption on Pd(111) and Pt(111) surfaces

The origin of the unique geometry for nitric oxide (NO) adsorption on Pd(111) and Pt(111) surfaces as well as the effect of temperature were studied by density functional theory calculations and ab initio molecular dynamics at finite temperature. We found that at low coverage, the adsorption geometry is determined by electronic interactions, depending sensitively on the adsorption sites and coverages, and the effect of temperature on geometries is significant. At coverage of 0.25 monolayer (ML), adsorbed NO at hollow sites prefer an upright configuration, while NO adsorbed at top sites prefer a tilting configuration. With increase in the coverage up to 0.50 ML, the enhanced steric repulsion lead to the tilting of hollow NO. We found that the tilting was enhanced by the thermal effects. At coverage of 0.75 ML with p(2 x 2)-3NO(fcc+hcp+top) structure, we found that there was no preferential orientation for tilted top NO. The interplay of the orbital hybridization, thermal effects, steric repulsion, and their effects on the adsorption geometries were highlighted at the end.

Prevalence, magnitude, and methods of rapid weight loss among judo competitors

ARTIOLI, G. G., B. GUALANO, E. FRANCHINI, F. B. SCAGLIUSI, M. TAKESIAN, M. FUCHS, and A. H. LANCHA. Prevalence, Magnitude, and Methods of Rapid Weight Loss among Judo Competitors. Med. Sci. Sports Exerc., Vol. 42, No. 3, pp. 436-442, 2010. Purpose: To identify the prevalence, magnitude, and methods of rapid weight loss among judo competitors. Methods: Athletes (607 males and 215 females; age = 19.3 +/- 5.3 yr, weight = 70 +/- 7.5 kg, height = 170.6 +/- 9.8 cm) completed a previously validated questionnaire developed to evaluate rapid weight loss in judo athletes, which provides a score. The higher the score obtained, the more aggressive the weight loss behaviors. Data were analyzed using descriptive statistics and frequency analyses. Mean scores obtained in the questionnaire were used to compare specific groups of athletes using, when appropriate, Mann-Whitney U-test or general linear model one-way ANOVA followed by Tamhane post hoc test. Results: Eighty-six percent of athletes reported that have already lost weight to compete. When heavyweights are excluded, this percentage rises to 89%. Most athletes reported reductions of up to 5% of body weight (mean +/- SD: 2.5 +/- 2.3%). The most weight ever lost was 2%-5%, whereas a great part of athletes reported reductions of 5%-10% (mean +/- SD: 6 +/- 4%). The number of reductions underwent in a season was 3 +/- 5. The reductions usually occurred within 7 +/- 7 d. Athletes began cutting weight at 12.6 +/- 6.1 yr. No significant differences were found in the score obtained by male versus female athletes as well as by athletes from different weight classes. Elite athletes scored significantly higher in the questionnaire than nonelite. Athletes who began cutting weight earlier also scored higher than those who began later. Conclusions: Rapid weight loss is highly prevalent in judo competitors. The level of aggressiveness in weight management behaviors seems to not be influenced by the gender or by the weight class, but it seems to be influenced by competitive level and by the age at which athletes began cutting weight.

Enhanced aeroelastic energy harvesting by exploiting combined nonlinearities: theory and experiment

Converting aeroelastic vibrations into electricity for low power generation has received growing attention over the past few years. In addition to potential applications for aerospace structures, the goal is to develop alternative and scalable configurations for wind energy harvesting to use in wireless electronic systems. This paper presents modeling and experiments of aeroelastic energy harvesting using piezoelectric transduction with a focus on exploiting combined nonlinearities. An airfoil with plunge and pitch degrees of freedom (DOF) is investigated. Piezoelectric coupling is introduced to the plunge DOF while nonlinearities are introduced through the pitch DOF. A state-space model is presented and employed for the simulations of the piezoaeroelastic generator. A two-state approximation to Theodorsen aerodynamics is used in order to determine the unsteady aerodynamic loads. Three case studies are presented. First the interaction between piezoelectric power generation and linear aeroelastic behavior of a typical section is investigated for a set of resistive loads. Model predictions are compared to experimental data obtained from the wind tunnel tests at the flutter boundary. In the second case study, free play nonlinearity is added to the pitch DOF and it is shown that nonlinear limit-cycle oscillations can be obtained not only above but also below the linear flutter speed. The experimental results are successfully predicted by the model simulations. Finally, the combination of cubic hardening stiffness and free play nonlinearities is considered in the pitch DOF. The nonlinear piezoaeroelastic response is investigated for different values of the nonlinear-to-linear stiffness ratio. The free play nonlinearity reduces the cut-in speed while the hardening stiffness helps in obtaining persistent oscillations of acceptable amplitude over a wider range of airflow speeds. Such nonlinearities can be introduced to aeroelastic energy harvesters (exploiting piezoelectric or other transduction mechanisms) for performance enhancement.