教学课程资源库 更多>>

关键词： Artificial intelligence   Computer science  

摘要： Probabilistic graphical models, e.g. Bayesian Networks, have been traditionally introduced to model and reason with uncertainty. A graph structure is crafted to capture knowledge of conditional independence relationships among random variables, which can enhance the computational complexity of reasoning. To generate such a graph, one sometimes has to provide vast and detailed knowledge about how variables interacts, which may not be readily available. In some cases, although a graph structure can be obtained from available knowledge, it can be too dense to be useful computationally. In this dissertation, we propose a new type of probabilistic graphical models called a Structured Bayesian network (SBN) that requires less detailed knowledge about conditional independences. The new model can also leverage other types of knowledge, including logical constraints and conditional independencies that are not visible in the graph structure. Using SBNs, different types of knowledge act in harmony to facilitate reasoning and learning from a stochastic world. We study SBNs across the dimensions of modeling, inference and learning. We also demonstrate some of their applications in the domain of traffic modeling.

关键词： Computer engineering   Computer science  

摘要： The exponentially increasing number of internet of things (IoT) devices and the data generated by these devices introduce the spectrum crisis at the already crowded ISM 2.4 GHz band. On one hand, there is no efficient solution to coordinate a large amount of heterogeneous IoT devices that have different communication technologies (e.g., WiFi, ZigBee, and Bluetooth). On the other hand, the crowded wireless traffic potentially introduces opportunities for conducting radio frequency (RF) sensing using pervasively available signals generated by heterogeneous IoT devices. To address these issues, we propose a multi-layer heterogeneous networking system to enable seamless communication and RF sensing among IoT devices. This networking system i) connects IoT devices with different communication protocols by utilizing their unique physical-layer or link-layer features on overlapped frequency band (e.g., ISM 2.4 GHz band); and ii) utilizes the heterogeneous wireless traffic for non-intrusive human activity recognition and monitoring, which provides data support for the emerging smart health and smart home applications. This system aims to significantly increase the throughput and spectrum utilization for wireless communication while keeping the low cost (in terms of energy and deployment) of IoT networks. Furthermore, the system provides sensing capability by using existing RF signals in the air.

关键词： Computer science  

摘要： As the ubiquity of financial technology services and applications continues to grow around the world, so does their utilization from businesses and consumers alike. This rapid growth poses a structural problem to the architecture upon which many of these applications were originally developed. This thesis highlights the operational characteristics and benefits of employing the microservice architectural style for these kinds of applications. It also details the necessity for high-quality software that adheres to the standards set forth by the microservice architectural style and its application to the financial technology (fintech) industry. A microservice-based application has also been developed and tested to serve as a living example of the power and flexibility of utilizing this approach. This thesis aims to add context and contribution to the field of software engineering with a specific focus in fintech SaaS application and the potential for their improvement through careful architectural consideration.

关键词： Computer Science  

摘要： Researchers have identified several quality metrics to predict defects, relying on different information however, these approaches lack metrics to estimate the effort of program understandability of system artifacts. Code that is understandable is often considered more maintainable. Programmers must have mental capacity for dealing with large workloads for short periods of time and must have cognitive mechanisms for locating the code relevant to a particular feature or variable. Program slicing is a technique that programmers use when scanning code for comprehension. In this dissertation, novel metrics to compute the cognitive complexity based on program slicing are introduced. These metrics help identify code that is more likely to have defects due to being challenging to comprehension. The metrics include such measures as the total number of slices in a file, the size, the average number of identifiers, and the average spatial distance of a slice. A scalable lightweight slicing tool is used to compute the necessary slicing data. A thorough empirical investigation into how cognitive complexity correlates with and predicts defects in the version histories of 10 datasets of 7 open source systems is performed. The results show that the increase of cognitive complexity significantly increases the number of defects in 94% of the cases. In a comparison study to metrics that have been shown to correlate with understandability and with defects, the addition of cognitive complexity metrics shows better prediction by up to 14% in F1, 16% in AUC, and 35% in R2.

关键词： Computer science  

摘要： Cyber-attacks, also referred to as network attacks, are becoming ubiquitous and affect all aspects of an organization, which is one of the most strategic risks as indicated by the World Economic Forum. There are several ways to detect these attacks, but these methods need continual updates and expansion as attackers increase the sophistication of attacks. According to a Cisco’s 2014’s threat intelligence report, 100% of the business networks analyzed had traffic going to websites that host malware, and 96% of the networks reviewed showed traffic to hijacked servers. The pervasiveness of malicious traffic indicates that organizations need to monitor network traffic closely and develop an effective solution to detect network traffic behaviors. This paper uses a machine learning model, Decision Tree, to analyze and apply classification to two network traffic datasets to identify network attacks and discover the attack patterns and types to predict future attacks. One dataset is from open source Security Repo, a security data samples repository; the second dataset is a company proprietary belong to a commercial bank. A Support Vector Machine model was used to validate the analysis accuracy. The results show, for both datasets, LinearSVC prediction accuracy is slightly higher than Decision Tree prediction accuracy. But Decision Tree models have higher True Negative Rate (TNR) in both datasets than LinearSVC models. The TNR indicates that the Decision Tree model returned more of the relevant results than the LinearSVC model; Decision Tree performed better prediction than LinearSVC in both datasets.

关键词： Training   Predictive models   Semisupervised learning   Data models   Reliability   Entropy   Computer science   Semi-supervised classification   deep ensemble   consistency regularization   landmark learning  

摘要： Using an ensemble of neural networks with consistency regularization is effective for improving performance and stability of deep learning, compared to the case of a single network. In this paper, we present a semi-supervised Deep Coupled Ensemble (DCE) model, which contributes to ensemble learning and classification landmark exploration for better locating the final decision boundaries in the learnt latent space. First, multiple complementary consistency regularizations are integrated into our DCE model to enable the ensemble members to learn from each other and themselves, such that training experience from different sources can be shared and utilized during training. Second, in view of the possibility of producing incorrect predictions on a number of difficult instances, we adopt class-wise mean feature matching to explore important unlabeled instances as classification landmarks, on which the model predictions are more reliable. Minimizing the weighted conditional entropy on unlabeled data is able to force the final decision boundaries to move away from important training data points, which facilitates semi-supervised learning. Ensemble members could eventually have similar performance due to consistency regularization, and thus only one of these members is needed during the test stage, such that the efficiency of our model is the same as the non-ensemble case. Extensive experimental results demonstrate the superiority of our proposed DCE model over existing state-of-the-art semi-supervised learning methods.

关键词： Artificial intelligence   Recommender Systems   Popularity Bias   Information science   Computer science   Personalization   Multi-stakeholder Recommendation   Long-tail   Fairness  

摘要： Traditionally, especially in academic research in recommender systems, the focus has been solely on the satisfaction of the end-user. While user satisfaction has, indeed, been associated with the success of the business, it is not the only factor. In many recommendation domains, there are other stakeholders whose needs should be taken into account in the recommendation generation and evaluation. In this dissertation, I describe the notion of multi-stakeholder recommendation. In particular, I study one of the most important challenges in recommendation research, popularity bias, from a multi-stakeholder perspective since, as I show later in this dissertation, it impacts different stakeholders in a recommender system. Popularity bias is a well-known phenomenon in recommender systems where popular items are recommended even more frequently than their popularity would warrant, amplifying long-tail effects already present in many recommendation domains. Prior research has examined various approaches for mitigating popularity bias and enhancing the recommendation of long-tail items overall. The effectiveness of these approaches, however, has not been assessed in multi-stakeholder environments. In this dissertation, I study the impact of popularity bias in recommender systems from a multi-stakeholder perspective. In addition, I propose several algorithms each approaching the popularity bias mitigation from a different angle and compare their performances using several metrics with some other state-of-the-art approaches in the literature. I show that, often, the standard evaluation measures of popularity bias mitigation in the literature do not reflect the real picture of an algorithm's performance when it is evaluated from a multi-stakeholder point of view.

关键词： Mathematics   Artificial intelligence   Computer science  

摘要： This goal of this thesis is to design and implement a light weighted AI for playing Gomoku with high level intelligence. Our work is built upon an innovative algebraic monomial theory to help assess values for each possible move and estimate chances for the AI to win at each move. With the help of the monomial theory, we are able to convert winning configurations into monomials of variables that represent the underlying board positions. In the existing approaches to building an AI for playing Gomoku, one common challenge is about how to represent the present configuration of the game along with the history of the moves of the two players. Compared with the usual 2D matrix of the board positions, our monomials can make the AI easily understand the current state and the history of the game, and they also allow the AI to compute the potential values for future moves from the current state and the history of moves made by the players. In addition, when we adopt the Monte Carlo Tree Search to probe for a possible winning strategy for the AI, those monomials help reduce the search space, in addition to help estimate rates for exploration of the historical moves and exploitation of the future moves. Based on the proposed algebraic monomial theory, we have implemented a lightweight powerful AI that is capable of playing Gomoku at highly competitive level. At this stage, our AI can win top rated AIs (up to top 7) from the most recent Gomocup rating.

关键词： Measurement   Pattern matching   Proteins   Computer science   Time series analysis   Internet   Bioinformatics   String metrics   generalized string similarity framework   edit distance   Jaccard distance  

摘要： In the last few years, we have assisted in a great increase of the usage of strings in the most disparate areas. In the meantime, the development of the Internet has brought the necessity of managing strings from very different contexts and possibly using different alphabets. This issue is not addressed by the numerous string comparison metrics previously proposed in the literature. In this paper, we aim at providing a contribution in this context. In fact, first we propose an approach to measure the similarity of strings based on different alphabets. Then we show that our approach can be specifically adapted to several classic string comparison metrics and that each specialization can lead to addressing completely different issues.

关键词： Cognitive psychology   Information science   Computer science  

摘要： Decisionmakers in business, government, and academia face an increasingly complex information environment, exacerbated by the limited cognitive processing capability of the human mind. Often, practitioners compensate for this decision-making challenge by synthesizing complex data into more comprehensible graphical representations. Unfortunately, the research into how data visualizations support complex cognitive activities such as decision-making is limited. This study explores constructive-interaction visualizations through a series of crowdsourced, randomized controlled experiments designed to determine if such affordances positively affect decision-making. It contributes to the emerging field of constructive-interactive visualization by uniquely integrating an exploration into its effects on higher-order cognitive processes. This study demonstrated that unlike previous research into lower-order cognitive functions such as perception, higher-order cognitive functions such as decision-making require deeper, more focused thought processes to make more accurate decisions. It provided additional, albeit inconclusive, evidence that interaction generally may aid decisionmakers in overcoming heuristics and biases that weaken the decision-making process. This study recommends expanded investigations into the ways constructive interaction supports higher-order cognitive functions, the development of a comprehensive framework for constructive interaction, and the integration of design science into the development of visualization artifacts.