Robots in the Classroom

The C* Conference on Computer Science and Software Engineering (C3S2E) was established in 2008. C3S2E 2010 is the third annual meeting in the C3S2E conference series and is held in Montréal, Québec, Canada, during May 19-20, 2010. The conference aims to address the need of the academic community in computer science and software engineering. It brings together academics, government and industry professionals to discuss recent progress and challenges in a broad range of areas in the field of computer science and software engineering such as algorithms and theory, artificial intelligence, bioinformatics, computer architecture, computer graphics, computer security, computer vision, data mining, databases, image processing, information retrieval, machine learning, as well as software engineering. C3S2E 2010 also serves as a platform for theoreticians and practitioners to exchange their original research ideas on academic or application aspects of computing and engineering of software systems, present their new findings on theoretical or practical aspects of computer science and software engineering, share their experiences on integrating new technologies into products and applications, discuss their work on applying computing and software engineering to real-life situations, describe their development and operations of challenging computing related systems, and identify unsolved challenges. Although it is the third time C3S2E is held, this comprehensive conference has grown from a Canadian conference into an international conference. For C3S2E 2010, we continue to keep our tradition for inviting international experts in various areas of computer science and software engineering to join our team of international program committee. As a result, our Program Committee consists of professionals from different parts of the world including Australia, Canada, Germany, Hong Kong, Italy, Japan, Portugal, South Korea, Turkey, UAE, and UK. This committee has done an excellent job in finishing the single-blind review and on-line double-blind debate processes in record time. On average, about 90% of reviews were submitted by the due date and about 3.67 reviews were received for each paper. In terms of submissions, we have attracted many high-quality papers submitted by authors globally. The paper selection process was thorough and competitive. On average, each paper was refereed by at least three reviewers, and about 18% of submissions were accepted as full research papers. This year, we have an intensive program that spans over two days. In addition to research papers and student posters, we also include position papers, which allow authors to give their opinions and vision on exploring new concepts and research directions.

Probabilistic graphical models (PGM) are one of the rich frameworks. These models are used over complex domains for coding probability distributions. The joint distributions interact with each other over large numbers of random variables and are the combination of statistics and computer science. These concepts are dependent on theories such as probability theory, graph algorithms, machine learning, which make a basic tool in devising many machine learning problems. These are the origin for the contemporary methods in an extensive range of applications. These applications range as medical diagnosis, image understanding, speech recognition, natural language processing, etc. Graphical models are one of dominant tools for handling image processing applications. On the other hand, the volume of image data gives rise to a problem. The representation of all possible graphical model node variables with that of discrete states heads to the number of states for the model. This leads to interpretation computationally obstinate. Many projects involve a human intervention or an automated system to obtain the consensus established on existing information. The PGM, discussed in this chapter, offers a variety of approaches. The approach is based on models and allows interpretable models to be built which then is employed by reasoning algorithms. These models are also studied significantly from data and allow the approaches for cases where the model is manually built. Most real-world applications are of uncertain data which makes a model building more challenging. This chapter emphasizes on PGM where the uncertainty of data is obvious. PGM provides models that are more realistic. These are extended from Bayesian networks, undirected Markov networks, discrete and continuous models, and extensions to deal with dynamical systems and relational data also. For each class of models, the chapter describes the fundamental bases: representation, inference, and learning. Finally, the chapter considers the decision making under the uncertainty of the data.

At the University of Tennessee at Martin (UTM), freshman computer science majors were being successful in their first quarter in computer science, but many majors were not completing the year sequence (Programming Concepts and Problem Solving I, II, and III). This study describes a freshman studies program that helps increase student retention in the first computer science course.Two populations were used in this study. The first population (Group A) consisted of 25 freshman computer science majors (16 men and 9 women) enrolled Fall Quarter 85 in freshman orientation General Studies GS1000 for one quarter hour credit. The second population (Group B) consisted of 17 freshman computer science majors (12 men and 5 women) enrolled the Fall Quarter 86 in freshman orientation General Studies GS1001 for two quarter hours credit. Both groups were placed in the first course in computer science using the cutting scores ACT English > 17 or 0.78 (ACT English) + 0.49 (UTM Math Placement) > 33. Both Groups A and B had an average ACT English score of 20 and an average UTM Math Placement score of 55. Each quarter during the freshman year, the computer science major enrolls in calculus (4 hrs.), Programming Concepts and Problem Solving (3 hrs.), a laboratory science (4 hrs.), English (3 hrs.), and a humanities course (3 hrs.).Group A was offered a freshman orientation program that involved a pass/fail one quarter hour credit GS1000 which met one hour each day during freshman studies week. Each section consisted of twenty-five heterogeneously grouped students. The course content was review of study skills. Group B was offered a freshman orientation program that involved a pass/fail two quarter hour credit GS1001 which met one hour each day during freshman studies week, plus one hour each week during the first quarter. Each section consisted of only students of a particular major (computer science) and was taught by a faculty member of that department. The topics of this course consisted of a review of study skills, discussions of student problems, special academic speakers, field trips to professional sites, cultural events, meetings with deans and department heads, and a follow-up on study skills. In the meeting before the final exam week, the faculty member dealt specifically with suggestions on preparing for finals.Groups A and B were enrolled in the same first computer science sequence (Programming Concepts and Problem Solving I, II, and III) their freshman year. Both groups had the same instructors, text (Introduction to Computing and Computer Science with Pascal by H. Walker), and course syllabus for this sequence. Both groups carried the same academic course load their freshman year. However, in Group A only 20% (5 out of 25) of the students completed the first computer science sequence with a grade of C or better, while in Group B 41% (7 out of 17) completed the same sequence with a grade of C or better. Student retention is achieved through the students' integration into the academic and social systems of the college by two factors: (1) faculty-student interaction and (2) peer-group interaction. Although the population of this study was small, I believe GS1001 provides one way of increasing student retention in computer science.

On 28-29 April 2000, the Board on Mathematical Sciences and the Computer Science and Telecommunications Board of the National Research Council cosponsored a workshop on the interface between the mathematical sciences and three areas of computer science: network traffic modeling, computer vision, and data mining. The workshop's goal was to identify some open research questions in the mathematical sciences whose solutions could contribute to important unsolved problems in those computer science disciplines. The 48 participants engaged in lively discussions and identified at least 36 research challenges of a mathematical, statistical, or probabilistic nature.

The 2025 3rd International Conference on Computer, Machine Learning and Artificial Intelligence (CMLAI 2025) had been held in San Francisco, USA during March 29-30, 2025. CMLAI 2025 is an international forum devoted specifically to facilitate synergies in research and development in the areas of computer science, machine learning, neural networks, algorithm design, and artificial intelligence. It provides the communication platform for those who are leading experts and scholars from around the world. The development of artificial intelligence and their corresponding technologies and industries has also been regarded as "national development strategy" for the world's leading countries and gradually developed into a new "trump card" in the game between countries. This conference was attended by over 60 researchers, scholars, academic and industry experts from many countries. During the two-day conference, 4 keynote speeches with topics on cloud computing, machine learning, neural networks, and artificial intelligence were delivered. And 24 oral and invited presentations were presented and discussed in-depth in 4 technical sessions, with interactive communication among the attendees. The conference successfully promoted exchanging new ideas and perspective discourse among the participants in their respective fields. CMLAI Organizing Committees San Francisco, USA

Neural network-based machine learning is capable of approximating functions\nin very high dimension with unprecedented efficiency and accuracy. This has\nopened up many exciting new possibilities, not just in traditional areas of\nartificial intelligence, but also in scientific computing and computational\nscience. At the same time, machine learning has also acquired the reputation of\nbeing a set of "black box" type of tricks, without fundamental principles. This\nhas been a real obstacle for making further progress in machine learning. In\nthis article, we try to address the following two very important questions: (1)\nHow machine learning has already impacted and will further impact computational\nmathematics, scientific computing and computational science? (2) How\ncomputational mathematics, particularly numerical analysis, {can} impact\nmachine learning? We describe some of the most important progress that has been\nmade on these issues. Our hope is to put things into a perspective that will\nhelp to integrate machine learning with computational mathematics.\n

Classification of the optical spectra of active galactic nuclei (AGN) into different types is currently based on features such as line widths and intensity ratios. Although well founded on AGN physics, this approach involves some degree of human oversight and cannot scale to large datasets. Machine learning (ML) tackles this classification problem in a fast and reproducible way, but is often (and not without reason) perceived as a black box. However, ML interpretability and are active research areas in computer science that are providing us with tools to mitigate this issue. We apply ML interpretability tools to a classifier trained to predict AGN types from spectra. Our goal is to demonstrate the use of such tools in this context, obtaining for the first time insight into an otherwise black box AGN classifier. In particular, we want to understand which parts of each spectrum most affect the predictions of our classifier, checking that the results make sense in the light of our theoretical expectations. We trained a support-vector machine on 3346 high-quality, low-redshift AGN spectra from SDSS DR15. We considered either two-class classification (type 1 versus 2) or multiclass (type 1 versus 2 versus intermediate-type). The spectra were previously and independently hand-labeled and divided into types 1 and 2, and intermediate-type (i.e., sources in which the Balmer line profile consists of a sharp narrow component superimposed on a broad component). We performed a train-validation-test split, tuning hyperparameters and independently measuring performance via a variety of metrics. On a selection of test-set spectra, we computed the gradient of the predicted class probability at a given spectrum. Regions of the spectrum were then color-coded based on the direction and the amount by which they influence the predicted class, effectively building a saliency map. We also visualized the high-dimensional space of AGN spectra using t-distributed stochastic neighbor embedding (t-SNE), showing where the spectra for which we computed a saliency map are located. Our best classifier reaches an F-score of 0.942 on our test set (with 0.948 precision and 0.936 recall). We computed saliency maps on all misclassified spectra in the test set and on a sample of randomly selected spectra. Regions that affect the predicted AGN type often coincide with physically relevant features, such as spectral lines. t-SNE visualization shows good separability of type 1 and type 2 spectra. Intermediate-type spectra either lie in-between, as expected, or appear mixed with type 2 spectra. Misclassified spectra are typically found among the latter. Some clustering structure is apparent among type 2 and intermediate-type spectra, though this may be an artifact. Saliency maps show why a given AGN type was predicted by our classifier resulting in a physical interpretation in terms of regions of the spectrum that affected its decision, making it no longer a black box. These regions coincide with those used by human experts, for example relevant spectral lines, and are even used in a similar way; the classifier effectively measures the width of a line by weighing its center and its tails oppositely.

It is tedious and error-prone to query search engines manually in order to accumulate a large body of factual information. Search engines retrieve and rank potentially relevant documents for human perusal, but do not extract facts, or fuse information from multiple documents. This paper introduces DynamicView, a Semantic Web application for researchers to query, browse and visualize distribution and evolution of research areas in computer science. Present and historical web pages of top 20 universities in USA and China are analyzed, and research areas of faculties in computer science are extracted automatically by segmentation based algorithm. Different ontologies of ACM and MST classification systems are combined by SKOS vocabularies, and the classification of research areas is learned from the ACM Digital Library. Query results including numbers of researchers and their locations are visualized in SVG map and animation. Interestingly, great differences of hot topics do exist between the two countries, and the number of researchers in certain areas changed greatly from the year 2000 to 2005.

At first sight, distributed computing and machine learning are two distant areas in computer science. However, there are many connections, for instance in the area of graphs, which are the focus of my talk. Distributed computing has studied distributed graph algorithms for many decades. Meanwhile in machine learning, graph neural networks are picking up steam. When it comes to dealing with graphical inputs, one can almost claim that graph neural networks are an application of distributed algorithms. I will introduce central concepts in learning such as underreaching and oversquashing, which have been known in the distributed computing community for decades, as local and congest models. In addition I am going to present some algorithmic insights, and a software framework that helps with explaining learning. Generally speaking, I would like to present a path to learning for those who are familiar with distributed message passing algorithms. This talk is based on a number of papers recently published at learning conferences such as ICML and NeurIPS, co-authored by Pál András Papp and Karolis Martinkus.

This paper surveys randomized parallel algorithms found in the literature for various problems in computer science. In particular we will demonstrate the power of randomization as a tool for parallelizing sequential algorithms and introduce the reader to some of the techniques employed in designing randomized parallel algorithms. We consider representative problems from the following areas of computer science and describe how randomized parallel algorithms for these problems have been obtained: 1)routing and sorting, 2)processor load balancing, 3)algebra, and 4)graph theory. Finally we discuss methods of derandomizing randomized parallel algorithms.KeywordsParallel AlgorithmDepth First SearchBoolean CircuitArithmetic ExpressionPram ModelThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

This paper describes a programming system used in the teaching of two Computer Science courses at The Pennsylvania State University (7). The courses are; an upper level undergraduate-beginning graduate level course in graph theory, and a graduate level course in the theory of graphs and networks. As can be ascertained from the description of these courses they are not primarily what would normally be called programming courses. However, since they are computer science courses they do stress the computational aspects of graph and network theory and analysis. Past experiences have indicated that if a computational problem of any reasonable complexity is assigned in class the majority of the students focus an inordinate amount of their energies to the programming problems, and consequently fail to grasp the significant features of the theoretical problems involved. In addition, since graph and network theory encompasses not only computer science, but engineering and physical, and to a lesser extent, the social sciences, many of these students would be unduly burdened by writing many rather complex programs. On the other hand, teaching the computational aspects of graph theory without doing some programming is equally ludicrous. Thus, a default type solution of assigning two or three problems from a narrow set of problems was rapidly becoming an undesirable solution to this dilema. It was at this point that the concept of a dynamically growing programming system, as described in this paper, began to formulate itself.

For the longest time computer systems have been modelled on the “informaton theoretic” view of communication characterized in Thomas 78 as follows: “Communication from system A to system B essentially involves the encoding of some internal state in system A into an external statement for transmission to system B. System B decodes the message and changes its internal state. Communication is considered “good” to the extent that there is an isomorphism between the internal states of the two systems after the message has been sent.” (p. 651). Recent advances in the design of expert and decision-support systems have brought about an increased awareness of the inadequacy of the encoding-decoding model for man-machine interaction. More appropriate is a functional model in which messages are passed in order to satisfy the goals of interaction. Thomas 78 proposes as an alternative, a game-theoretic model of communication. This paper argues that man-machine communication should be modelled on theories of human communication in natural language.Linguists and philosophers of language, in particular Searle 69 and Grice 75, have proposed functional theories of language use in which linguistic behaviour is characterized as a goal pursuing activity. This paper does not propose that natural language should be introduced as a universal vehicle for computer interaction but rather that the design of communicatively adequate computer systems should be modelled on the principles governing natural language use. The practice of modelling computer systems after paradigms of human cognitive behaviour has proved successful in other areas of computer science, in particular in computer vision. Research in natural language processing should hence not be viewed as a peripheral area in computer science merely aimed at providing the 'luxury' of allowing the casual user unrestricted means of interaction but as a prerequisite for the design of the next generation computer languages.

The publication provides some summary and methodological guidelines for topics, learning tasks, research tasks and projects related to colors. The authors present their own summary of the scientific knowledge of color (taxonomy) based on their experience in various areas of computer science (coding, programming, computer vision, structural combinatorics, user interfaces and usability) and their participation in interdisciplinary research and, in accordance with their pedagogical experience, provide possible projections of this knowledge in the teaching of informatics, IT, and computer science for different ages and experience of students. Examples of learning tasks and resources from already established textbooks are given, as well as new educational resources for color coding, color transformation, color schemes and palettes, index colors, heat maps on data, colors in user interfaces and design, emotions and colors, steganography, as well as educational environments in which to implement the tasks: programming environment, simulation environment for experiments, artificial intelligence. A special focus is on STEAM research tasks and projects related to color. Examples of new directions in the science of multimodal analysis are also demonstrated through multimodal tasks, namely parallel processing of text-sound-images related to color.

Artificial intelligence is a spectacular part of computer engineering that has earned a compelling diversion in the field of medical data classification due to its state-of-art algorithmic strength and learning capabilities. Machine Learning is a major sub-domain of artificial intelligence, where it has become one of most promising fields in computer science. In recent years, there is a large spectrum of healthcare and biomedical data has been growing intensely. Due to the huge labeled or unlabeled data, it is important to have a compact and robust machine learning solution for classification. Several optimizers have been deployed to improve the inclusive performance of machine learning models. The classification of machine learning models depend on several factors. This comprehensive review paper aims to insight the current stage of optimized machine learning success on medical data classification. Increasing number of unstructured medical data has been utilizing in machine learning algorithms to predict intuitions. But it is difficult to inherent immense intuition from those data. So machine learning researchers have utilized state-of-art optimizers and novel feature selection techniques to overcome and emend the performance accuracy. We have highlighted some recent literatures, which exhibits the robust impact of optimizers and feature selection on machine learning techniques on medical data characterization. On the other hand, a clean-cut introduction on machine learning and theoretical outlook of widely utilized optimization techniques like genetic algorithm, gray wolf optimization, and particle swarm optimization are discussed for initial understanding to the optimization techniques.

Situation is a computational abstraction that encapsulates human-centric contexts, which can be human-oriented behavioral contexts as well as the relevant environmental contexts. Its potential applications in computer science range from artificial intelligence, service computing, human-computer interaction, pervasive computing to software engineering and software systems. In this paper, we introduce a user-centric situation-aware Intelligent Online Learning Platform (IOLP) that aims to provide a personalized learning experience for those conducting online-learning activities. It joins the engineering techniques from Machine Learning, Intelligent Systems (IS) and Human-Computing Interaction (HCI) to offer highly personalized services to the users. It serves as an experimental subject for, as well as a real-world implementation of situations. Under the strength of the abstraction of situation, various elements drawn from different areas in computer science are seamlessly integrated towards a user-centric Intelligent Online Learning Platform (IOLP). Our discussions on the IOLP are further supported by an online class instance it currently hosts featuring Python Programming.