Research In Theoretical Computer Science Research Articles

Online List Scheduling for Makespan Minimization

In the modern interactive computing era, computational problems such as scheduling, rout- ing, sequencing, and resource management are online in nature. In the online framework, at the outset, an algorithm receives and processes inputs one by one in order without the knowledge of future inputs, unlike in an offline framework, where an algorithm knows the entire input sequence at the beginning. The design of mathematical models and efficient al- gorithms for the online framework constitutes a challenging and competitive area of research in Theoretical Computer Science. The researchers have considered competitive analysis as a well-de ned and standard method to measure the performance of online algorithms.

Best Practice Leads to the Best Materials Informatics

Developing and employing best practice is an important stage in a maturing scientific discipline and is well overdue for the field of materials informatics, where a systematic approach to data science is needed to ensure the reliability and reproducibility of results. Developing and employing best practice is an important stage in a maturing scientific discipline and is well overdue for the field of materials informatics, where a systematic approach to data science is needed to ensure the reliability and reproducibility of results. Machine learning was developed in the field of computer science over the last 60 years as a way for computers to effectively program themselves.1Mayo H. Punchihewa H. Emile J. Morrison J. History of Machine Learning.https://www.doc.ic.ac.uk/∼jce317/history-machine-learning.htmlDate: 2018Google Scholar It was Arthur Samuel who first termed “Machine Learning” in 1952 to describe his research, which, combined with Donald Hebb’s model of brain cell interaction, later lead Frank Rosenblatt to create the first perceptron in 1957: the precursor to modern day neural networks and deep learning. To the field of materials science, having begun to use machine learning in earnest over the last ~15 years,2Rajan K. Combinatorial Materials Sciences: Experimental Strategies for Accelerated Knowledge Discovery.Annu. Rev. Mater. Res. 2008; 38: 299-322Crossref Scopus (101) Google Scholar it is a new and powerful way to navigate physical complexity. Machine learning allows us to retain as much information as possible, without needing to make pragmatic compromises and decide at the outset what to ignore, allowing our structure/property predictions to uncover themselves. Machine learning, however, is not a panacea to all problems, nor is it a magic bullet to extract answers to questions you have not yet thought to ask. Machine learning is a tool that performs a task, and although we (humans) might not be able to comprehend the complexity of our data directly, we still need to formulate our questions well. One cannot simply pour the data in and see what comes out. The recent democratization of machine learning methods is largely responsible for the rapid increase in materials informatics studies we see in the literature today and the troubling assumption that we can do just that. It is also responsible for the growing cohort of researchers that appear to associate an ability to download and install computational methods with the ability to use them correctly. These skills are not mutually inclusive and, unfortunately, there are a lot of misguided or incorrect machine learning studies of materials being published. Insufficient attention is being paid to the decades of research in theoretical computer science that lead to the development of these methods and the very specific types of datasets they were each intended for. This is not to say that we cannot leverage this rich reserve of algorithms, but care should always be taken when translating methods from one domain into another. This is intuitively obvious to material scientists when transposing methods from chemistry, physics, or engineering, and one should be just as vigilant when using methods from computational science. Creating and storing a large amount of scientific data is not the same as being able to do data science. The new editorial in Chemistry of Materials by Wang et al.3Wang A.Y.-T. Murdock R.J. Kauwe S.K. Oliynyk A.O. Gurlo A. Brgoch J. Persson K.A. Sparks T.D. Machine Learning for Materials Scientists: An Introductory Guide toward Best Practices.Chem. Mater. 2020; https://doi.org/10.1021/acs.chemmater.0c01907Crossref PubMed Scopus (32) Google Scholar brings this issue to the fore and parses the many steps in a materials informatics workflow that must be undertaken to ensure accuracy, reliability, and reproducibility. These steps are founded in statistics, software development, and scientific integrity and include a detailed description of lots of the boring bits (like data processing and feature extraction), as well as the exciting bits like model fitting and visualization. Although in many parts the editorial seems unnecessarily repetitive, this is not the case. This repetition serves to reinforce some of the most important data science principles to consider well before undertaking any machine learning. Much of materials informatics is in describing, cleaning, sorting, and reducing the data. Comparatively little time is spent generating models, and the majority of the time that is spent on machine learning involves hyper-parameter optimization and model validation. The same can be said for the emergent field of nanoinformatics.4Barnard A.S. Motevalli B. Parker A.J. Fischer J.M. Feigl C.A. Opletal G. Nanoinformatics, and the big challenges for the science of small things.Nanoscale. 2019; 11: 19190-19201Crossref PubMed Google Scholar The actual machine learning is only the tip of the data science iceberg (Figure 1). This timely addition to the materials informatics literature includes a handy interactive electronic notebook demonstrating an entire workflow to predict the heat capacity of solid inorganic compounds, but the same principles apply to any feature space (structures) and target labels (properties). This is a useful resource, and some researchers will find it tempting to simply swap out the dataset and reapply this workflow verbatim. Those that have carefully read and understood the editorial will, however, understand that it is just this sort of black-box “plug and play” approach to materials informatics that has led to such poor research being published; this is precisely why best practice needs to be established. Every dataset is different. Every dataset has nuances that require different processing methods. Every dataset has biases that must be addressed, or our results will be misleading (at best) or entirely wrong. Although the editorial focuses on the use of high-throughput computational data, this does not prevent the same protocols from being applied to high-throughput experimental data or any other sufficiently large and comprehensive database generated through computational or experimental studies. This being said, the advice provided in this editorial is very specific to materials science and is targeted toward beginners in the field. There is a range of other data science protocols that should be considered by advanced practitioners, such as data fusion methods and protocols to combine resources, data augmentation methods to expand a dataset, under-sampling or over-sampling to handle imbalance (particularly when splitting datasets to avoid data bleeding), and data processing methods to deal with systematic bias (such as selection bias, evaluation bias, or information bias). The latter is particularly important, given that machine learning amplifies bias. On a final note, it is important to highlight that best practice applies to all researchers, which means reviewers as well as authors. Inexperienced reviewers ignoring best practice, or requiring authors to unnecessarily include extra models that are poorly suited to the problem or dataset at hand, are equally responsible for the dissemination of “worst practice.” We need all members of the community doing their job well if we are to lay a strong foundation for the future of rational materials discovery and design.

Research in theoretical computer science

Theoretical computer science has been a vibrant part of computing research in India for the past 30 years. India has always had a strong mathematical tradition. One could also argue that in the 1980s and 1990s, theory offered a unique opportunity to keep up with international research in computing despite limited access to state-of-the-art hardware.

On Analyzing the Impact of Authors and Their Collaboration Patterns in the Major Computer Algorithms Research Conferences

An algorithm represents the computer-based road map for accomplishing a given, well-defined task. In general, computer algorithms form one of the core practical technologies of computer science. In principle, research in theoretical computer science focuses on the design and analysis of algorithms and data structures, as well as the study of fundamental barriers to efficient computation. The research community of this field has been significantly growing in the last decade. In general, in the computer science field, disseminating research outcomes is an integral part of a scholar’s professional career. The preference for conference publications is motivated by many arguments such as the high dynamism of the field requiring shorter turnaround time between submission; the prestige associated with publishing at highly-selective venues. We analyze the publications of three top-tier computer algorithms research conferences (SODA, ICALP and SOCG) over the last two decades (1995 - 2014). In particular, we report the list of the scholars who have the highest number of publications for each venue separately and when combined. We also analyze the co-authors relationships to understand their research collaborations. Finally, we study the increase in the number of publications and the size of the research community in the last two decades.

On Unknown Small Subsets and Implicit Measures: New Techniques for Parameterized Algorithms

Parameterized computation is a recently proposed alternative approach to dealing with NP-hard problems. Developing efficient parameterized algorithms has become a very active research area in the current research in theoretical computer science. In this paper, we investigate a number of new algorithmic techniques that were proposed and initiated by ourselves in our research in parameterized computation. The techniques have proved to be very useful and promising, and have led to improved parameterized algorithms for many well-known NP-hard problems.

A simple algorithm for 4-coloring 3-colorable planar graphs

Graph coloring for 3-colorable graphs receives very much attention by many researchers in theoretical computer science. Deciding 3-colorability of a graph is a well-known NP-complete problem. So far, the best known polynomial approximation algorithm achieves a factor of O ( n 0.2072 ) , and there is a strong evidence that there would be no polynomial time algorithm to color 3-colorable graphs using at most c colors for an absolute constant c . In this paper, we consider 3-colorable PLANAR graphs. The Four Color Theorem (4CT) (Appel and Haken (1977) [1], Appel et al. (1977) [2], Robertson et al. (1997) [14]) gives an O ( n 2 ) time algorithm to 4-color any planar graph. However the current known proof for the 4CT is computer assisted. In addition, the correctness of the proof is still lengthy and complicated. We give a very simple O ( n 2 ) algorithm to 4-color 3-colorable planar graphs. The correctness needs only a 2-page proof.

An algebraic approach for quantum computation

Quantum computation is one of the very promising new fields of the new millennium for basic research in theoretical computer science and computational mathematics. Quantum computers have stimulated...

The Dual Complexity Space as the Dual of a Normed Cone

In [M. Schellekens, The Smyth completion: A common foundation for denotational semantics and complexity analysis, in: Proc. MFPS 11, Electronic Notes in Theoretical Computer Science, vol. 1 (1995), 23 pages] M. Schellekens introduced the complexity (quasi-metric) space as a part of the research in Theoretical Computer Science and Topology, with applications to the complexity analysis of algorithms. Later on, S. Romaguera and M. Schellekens ([S. Romaguera, M. Schellekens, Quasi-metric properties of complexity spaces, Topology Appl. 98 (1999), 311–322]) introduced the so-called dual complexity (quasi-metric) space and established several quasi-metric properties of the complexity space via the analysis of th e dual. These authors also proved in [S. Romaguera, M. Schellekens, Duality and quasi-normability for complexity spaces, Appl. Gen. Topology 3 (2002), 91–112] that actually the dual complexity space C∗ can be modeled as a norm-weightable cone whose induced quasi-metric is Smyth complete. This fact suggests the existence of deep connections between a general theory of (dual) complexity spaces and Asymmetric Functional Analysis. These connections have been recently explored in [L.M. Garcá-Raffi, S. Romaguera, E.A. Sánchez-Pérez, Sequence spaces and asymmetric norms in the theory of compuational complexity, Math. Comput. Model 36 (2002), 1–11], [L.M. Garcá-Raffi, S. Romaguera, E.A. Sánchez-Pérez, The supremum asymmetric norm on sequence spaces: a general framework to measure complexity distances, in: Proceedings of the Second Irish Conference on the Mathematical Foundations of Computer Science and Information Technology (MFCSIT 2002), Galway, Ireland, July 2002; Electronic Notes in Theoret. Comput. Sci. 74 (2003), URL: http://www.elsevier.nl/locate/entcs/volume74.htm 12 pages] and [M. O'Keefe, S. Romaguera, M. Schellekens, Norm-weightable Riesz spaces and the dual complexity space, in: Proceedings of the Second Irish Conference on the Math ematical Foundations of Computer Science and Information Technology (MFCSIT 2002), Galway, Ireland, July 2002; Electronic Notes in Theoret. Comput. Sci. 74 (2003), URL: http://www.elsevier.nl/locate/entcs/volume74.htm 17 pages]. In particular, it was proved in [L.M. Garcá-Raffi, S. Romaguera, E.A. Sánchez-Pérez, Sequence spaces and asymmetric norms in the theory of compuational complexity, Math. Comput. Model 36 (2002), 1–11] that the so-called dual p-complexity space Cp∗, with 1⩽p<∞, is isometrically isomorphic to the positive cone of the classical Banach space lp. The space C1∗ is exactly the dual complexity space, and thus it is isometrically isomorphic to the positive cone of the Banach space l1 of all absolutely summable real sequences. Here, we continue the analysis of the structure of the dual complexity space C∗. We show that it is the dual space of the positive c one of the Banach space c0 of all real sequences converging to zero, and that its dual space is the positive cone of the Banach space l∞ of all bounded real sequences. Furthermore, the dual space of Cp∗, 1<p<∞, is Cq∗ where 1/p+1/q=1. These results extend to this setting well-known theorems of the classical theory of Functional Analysis.

Investigations on the Dual Calculus

The Dual Calculus, proposed recently by Wadler, is the outcome of two distinct lines of research in theoretical computer science: (A) Efforts to extend the Curry–Howard isomorphism, established between the simply-typed lambda calculus and intuitionistic logic, to classical logic. (B) Efforts to establish the tacit conjecture that call-by-value (CBV) reduction in lambda calculus is dual to call-by-name (CBN) reduction. This paper initially investigates relations of the Dual Calculus to other calculi, namely the simply-typed lambda calculus and the Symmetric lambda calculus. Moreover, Church–Rosser and Strong Normalization properties are proven for the calculus’ CBV reduction relation. Finally, extensions of the calculus to second-order types are briefly introduced.

General Recursion via Coinductive Types

A fertile field of research in theoretical computer science investigates the representation of general recursive functions in intensional type theories. Among the most successful approaches are: the use of wellfounded relations, implementation of operational semantics, formalization of domain theory, and inductive definition of domain predicates. Here, a different solution is proposed: exploiting coinductive types to model infinite computations. To every type A we associate a type of partial elements Partial(A), coinductively generated by two constructors: the first, return(a) just returns an element a:A; the second, step(x), adds a computation step to a recursive element x:Partial(A). We show how this simple device is sufficient to formalize all recursive functions between two given types. It allows the definition of fixed points of finitary, that is, continuous, operators. We will compare this approach to different ones from the literature. Finally, we mention that the formalization, with appropriate structural maps, defines a strong monad.

Data Streams: Algorithms and Applications

In the data stream scenario, input arrives very rapidly and there is limited memory to store the input. Algorithms have to work with one or few passes over the data, space less than linear in the input size or time significantly less than the input size. In the past few years, a new theory has emerged for reasoning about algorithms that work within these constraints on space, time, and number of passes. Some of the methods rely on metric embeddings, pseudo-random computations, sparse approximation theory and communication complexity. The applications for this scenario include IP network traffic analysis, mining text message streams and processing massive data sets in general. Researchers in Theoretical Computer Science, Databases, IP Networking and Computer Systems are working on the data stream challenges. This article is an overview and survey of data stream algorithmics and is an updated version of [1].

Preface: Volume 45

ForewordThis volume contains the Proceedings of MFPS 17, the Seventeenth Conference on the Mathematical Foundations of Programming Semantics. The conference was hosted by the Basic Research in Computer Science and Aarhus University, Aarhus, Denmark, and took place from May 23 through May 26, 2001.MFPS 17 featured invited talks by seven renowned scholars: •Olivier Danvy Aarhus•Joshua Guttman Mitre Corporation•Neil Jones DIKU•Kim Larsen Aalborg•Prakash Panangaden McGill•Jan Rutten CWI•Glynn Winskel CambridgeThere also were three special sessions: 1.A session honoring Neil Jones (DIKU) organized by Olivier Danvy (Aarhus) and David Schmidt (Kansas State University). The session began with Olivier Danvy's invited address, and included talks by Radhia Cousot (Ecole Polytechnique, Paris), John Hannan (Penn State), John Hughes (Chalmers), David Schmidt and Peter Sestoft (ITU and Royal Veterinary and Agricultural University).2.A session on model checking organized by Gavin Lowe (Oxford). The session began with Kim Parsen's invited address, and included talks by Jose Desharnais (Laval), Michael Huth (Kansas State and Imperial College), Henrik Jensen (BRICS), Marta Kwiatkowska (Birmingham) and Gavin Lowe.3.A session on security organized by Catherine Meadows (NRL). The session began with Joshua Guttman's invited address, and included talks by Andrew Gordon (Microsoft, Cambridge) and Alan Jeffrey (DePaul), Gavin Lowe, Thomas Jensen (INRIA), Dusko Pavlovic (Kestrel), Catherine Meadows and Andre Scedrov (Penn).The remainder of the program was made up of papers selected by the Program Committee from those submitted in response to the Call for Papers that was circulated in Fall, 2000. The Program Committee was co-chaired by Stephen Brookes (CMU) and Michael Mislove (Tulane) and included Lars Birkedal (ITU), Rance Cleaveland (SUNY, Stony Brook), Marcelo Fiore (Cambridge), Matthew Hennessy (Sussex), Alan Jeffrey (DePaul), Achim Jung (Birmingham), Gavin Lowe (Oxford), Catherine Meadows (NRL), Peter O'Hearn (QMW), Suysan Older (Syracuse), Dusko Pavlovic (Kestrel), Uday Reddy (Birmingham), Giuseppe Rosolini (Genoa), Davide Sangiori (INRIA) and Andre Scedrov (Penn). The Organizers of the MFPS series are Stephen Brookes, Michael Main (Colorado), Austin Melton (Kent State), Michael Mislove and David Schmidt. The Local Arrangements for the meeting were overseen by Olivier Danvy and Andrzej Filinski, both of BRICS and Aarhus. We are grateful to these colleagues for having so efficiently overseen the local arrangements. We also are happy to acknowledge the exceptional efforts of Karen Kjer Müller, who literally anticipated almost all of the needs of participants, and saw to it that every aspect of the meeting went off without any problems.Finally, we are extremely grateful for the financial support provided by BRICS and by the US Office of Naval Research, without which the meeting could not have taken place. We are particularly grateful to Dr. Ralph Wachter for his continuing interest in the MFPS series and the support his program in Software Systems has provided the series.Stephen Brookes and Michael Mislove, Guest Editors DedicationThe Organizers of the MFPS series are pleased to dedicate these Proceedings to Neil Jones for his continuing inspiration to the students and researchers in theoretical computer science. Neil has been a regular participant in the MFPS series, dating from his invited address at MFPS III in 1987. MFPS appreciates the continued inspiration that his teaching and research have provided and that his addresses to the MFPS community have so clearly elucidated.

Philippe Flajolet's Research in Analysis of Algorithms and Combinatorics

Philippe Flajolet's research in theoretical computer science spans over more than 20 years. He made lasting contributions to the analysis of algorithms and analytic combinatorics. Among many of his results we mention here some in such diversified topics as enumeration, number theory, formal languages, continued fractions, automatic analysis of algorithms, Mellin transform, digital sums, recurrences, trees, random generation of combinatorial objects, random graphs and mappings, polynomial factorization, communications, codes, graphics, etc. This gives only a small snapshot of his work, and we encourage the reader to visit Flajolet's homepage http://www-rocq.inria.fr/algo/flajolet/index.html for a fuller account. The bibliography was taken from his homepage and may not be totally complete, although we added a few items. Our paper was written without giving any prior notice to Philippe Flajolet. It reflects the view of the authors and any misunderstandings and shortcomings should be put on their account.