教学课程资源库 更多>>

关键词： Industrial engineering   Computer science   Artificial intelligence  

摘要： Student retention is a critical metric for many universities whose intention is to support student success. The goal of this thesis is to create retention models utilizing machine learning (ML) techniques. The factors explored in this research include only those known during the admissions process. These models have two goals: first, to correctly predict as many non-returning students as possible, while minimizing the number of students who are falsely predicted as non-returning. Next, to identify important features in student retention and provide a practical explanation for a student's decision to no longer persist. The models are then used to provide outreach to students that need more support. The findings of this research indicate that the current top performing model is Adaboost which is able to successfully predict non-returning students with an accuracy of 54 percent.

关键词： Computer engineering   Computer science   Engineering  

摘要： Cyber-Physical Systems (CPS) are physical entities whose operations are monitored, coordinated, and controlled by a computing and communication core. These systems are highly heterogeneous and complex. Their numerous components and cross domain complexity make attacks easy to propagate and security difficult to implement. Consequently, to secure these systems, they need to be built in a systematic and holistic way, where security is an integral part of the development lifecycle and not just an activity after development. These systems present a multitude of implementation details in their component units, so it is fundamental to use abstraction in the analysis and construction of their architecture. In particular, we can apply abstraction through the use of patterns. Pattern-based architectural modeling is a powerful way to describe the system and analyze its security and the other non-functional aspects. Patterns also have the potential to unify the design of their computational, communication, and control aspects. Architectural modeling can be performed through UML diagrams to show the interactions and dependencies between different components and its stakeholders. Also, it can be used to analyze security threats and describe the possible countermeasures to mitigate these threats. An important type of CPS is a maritime container terminal, a facility where cargo containers are transported between ships and land vehicles; for example, trains or trucks, for onward transportation, and vice versa. Every cargo port performs four basic functions: receiving, storing, staging and loading for both, import and export containers. We present here a set of patterns that describe the elements and functions of a cargo port system, and a Reference Architecture (RA) built using these patterns. We analyze and systematically enumerate the possible security threats to a container terminal in a cargo port using activity diagrams derived from selected use cases of the system. We describ

关键词： Design   Physiology   Biology   Computer science   Mathematics   Physics  

关键词： Computer science   Artificial intelligence  

摘要： A modern urban infrastructure no longer operates in isolation but instead leverages the latest technologies to collect, process, and distribute aggregated knowledge to improve the quality of the provided services and promote the efficiency of resource consumption. However, the ambiguity of ever-evolving cyber threats and their debilitating consequences introduce new barriers for decision-makers. Numerous techniques have been proposed to address the cyber misdemeanors against such critical realms and increase the accuracy of attack inference; however, they remain limited to detection algorithms omitting attack attribution and impact interpretation. The lack of the latter prompts the transition of these methods to operation difficult to impossible. In this dissertation, we first investigate the threat landscape of smart cities, survey and reveal the progress in data-driven methods for situational awareness and evaluate their effectiveness when addressing various cyber threats. Further, we propose an approach that integrates machine learning, the theory of belief functions, and dynamic visualization to complement available attack inference for ICS deployed in the realm of smart cities. Our framework offers an extensive scope of knowledge as opposed to solely evident indicators of malicious activity. It gives the cyber operators and digital investigators an effective tool to dynamically and visually interact, explore and analyze heterogeneous, complex data, and provide rich context information. Such an approach is envisioned to facilitate the cyber incident interpretation and support a timely evidence-based decision-making process. Empirical evaluations using data collected in a testbed representing a small-scale water treatment plant exceeded state-of-the-art performance. The results also show that the proposed approach identifies the exploited ICS assets with more than 8-15% accuracy improvement over currently available works. We postulate and stress that such propose

关键词： Computer science   Biology  

关键词： Islands   Airports   Algebra   Polynomials   Algorithms   Mathematics   Computer science   Transportation  

摘要： In the study of algebraic varieties over number fields or finite fields, there are many properties of these varieties that can be difficult to compute. This thesis discusses variants of the following three computational problems: (a) to compute an isogeny between two given supersingular elliptic curves; (b) to determine whether a zero-cycle on a product of elliptic curves is rationally equivalent to zero, and if so, to compute a rational equivalence; (c) to compute the space of cuspidal modular forms over a function field of genus greater than one. For each of these problems, we present algorithms that allow the problem to be solved in certain special cases, and prove results about these cases.

关键词： Computer science   Artificial intelligence  

摘要： A solar flare is a sudden flash of increased brightness usually observed near the surface of the sun and in close proximity to a sunspot group. Powerful flares are often accompanied by a phenomenon called coronal mass ejection. A coronal mass ejection is a high energy burst of plasma from the solar corona. These high energy plasmas can significantly affect technologies such as satellites in orbit which can severely impact human activity. Forecasting solar flares can therefore be of vital importance. These flares usually occur in regions of strong magnetic field called active regions which can be observed using a 2D pictorial representation called a magnetogram. Forecasting the solar flares using deep learning/ machine learning has become very prevalent in recent years. This is usually done using classification-based techniques on magnetograms, categorizing them into either flaring or non-flaring classes in the next k hours. These models commonly use a feature set extracted using the physics of the active region. Another feature set for forecasting solar flares uses topological data analysis on the magnetogram to extract shape-based features. In this thesis, I look into models that improve the forecasting accuracy using both these sets of features. As flaring of an active region is linked with the nature and evolution of the active region, a model that captures the temporal evolution of the active region has the potential to perform better. To achieve this, I will build a model that uses a temporal sequence of these features to forecast whether a flare occurs in the next 24 hours. Besides this, I will improve the current feature set by removing the redundancies. Too many features can lead to overfitting of the deep learning model and result in lower performance. By removing the redundancies, I will construct a smaller feature set that performs better than the original feature set. Along with this, I will also explore better approaches to constructing the topological

关键词： Control theory   Work stations   Software   Collaboration   Artificial intelligence   Planning   Logic   Robots   Design   Engineering   Computer science   Robotics  

摘要： In the automation systems of tomorrow, it is likely that the devices included have various degrees of autonomy, and include advanced algorithms for perception and control. Human operators will be expected to work together with collaborative robots as well as with roaming robots for material handling. The volatile nature of the environment of such intelligent automation systems lead to an enormous amount of possible situations that can arise and which need to be suitably handled. This complexity makes development of control systems for intelligent automation systems difficult using traditional methods. As an alternative, this thesis presents a model-based control framework, which uses a combination of formal specification and automated planning. The proposed framework allows for defining the intentions of the automation system on a high level, which enables decisions that influence when things should occur to be modeled using logical constraints, rather than programming. To achieve a modular framework, low level, reusable, resource models are composed by 1) formal specification to ensure safety and 2) applying an abstraction called an operation, which couples the reusable resources to the intentions of the system. By planning also the resources' detailed actions, the operations can, when possible, be completed regardless of the resources' current state. This eases error-recovery, as resources do not have to be reset when an error occurs. Additionally, the thesis proposes an iterative and interactive workflow for integrating the proposed model-based control framework into a virtual preparation process, using computer-based simulation as a tool for validating formal specifications. The control framework allows for adding new constraints to a running system, enabling an efficient and interactive preparation process. The framework has been applied to a use case from final assembly, which features human-robot collaboration. Experimental results on the ability to handle un

关键词： Computer science   Engineering  

摘要： The field of Digital Histopathology has seen many significant advancements in the past decade, and this is primarily because of the developments that have been seen in Artificial Intelligence, pertaining to the segmentation and classification of cell types in Whole Slide Images (WSI) of tissues. The focus of these developments has been on neoplastic or cancerous cell types in adults as the primary area of research. In this work, we propose a transfer learning approach to improve the classification of inflammatory cells for pediatric patients, as it is an emerging research area that is gaining momentum in the medical field. We conduct a review to identify the existing models in this field and propose a novel approach to improve the model's performance through the usage of transfer learning. We chose the PanNuke dataset, as it provides labeled tissue samples that have been expertly reviewed and have been shown to encompass real-world variations that would be experienced in a real-world setting, making it a more practical choice for training the model. We showcase that our SqueezeNet model developed on *** archives state-of-the-art performance on the PanNuke dataset, with an overall weighted average F1-Score of 89.2% on the binary classifier (between inflammatory cells and all other types of cells.) It also achieves 79.3% on the multi-class classifier (between all five cell types included in the PanNuke dataset.) We also show that our score achieves an improvement of 41% in the weighted average F1-Score against the previous state-of-the-art Hover-Net model for the Neoplastic, Epithelial, Inflammatory, and Connective cell types. Furthermore, we demonstrate achieving state-of-the-art sensitivity and precision scores in all trained cell types for the classification task for individual class performance, and the overall model weighted average.

关键词： Mechanical engineering   Computational chemistry   Deep learning   Artificial intelligence   Graph representations   Eigen values   Hybridization   Molecular weight   Chemical bonds   Algorithms   Computer science   Design  

摘要： Recent advances in artificial intelligence have propelled the development of innovative computational materials modeling and design techniques. Generative deep learning models have been used for molecular representation, discovery, and design. In this work, we assess the predictive capabilities of a molecular generative model developed based on variational inference and graph theory in the small data regime. Physical constraints that encourage energetically stable molecules are proposed. The encoding network is based on the scattering transform with adaptive spectral filters to allow for better generalization of the model. The decoding network is a one-shot graph generative model that conditions atom types on molecular topology. A Bayesian formalism is considered to capture uncertainties in the predictive estimates of molecular properties. The model's performance is evaluated by generating molecules with desired target properties.