教学课程资源库 更多>>

关键词： Computer science education   K-12 teacher professional development   STEM education   Systems theory   Teacher education  

摘要： Educational reform in the United States is often focused on upgrading school curricula in response to evolution in science, technology, and society. Wyoming is no exception. A 2001Wyoming Supreme Court decision granted constitutional protection to every K-12 student in the state for a “fair, complete, and equal education, appropriate for the times.” Based on that decision, in 2015 Governor Matt Mead initiated an executive action to make computer science education available in every Wyoming school. Such a new mandate caught many Wyoming schools off guard in that their in-service teachers were required to obtain education and certification to provide computer science education. To support the state’s schools, several of Wyoming’s institutions of higher education began delivering professional development to teachers needing qualifications to teach modern computer science. The research presented in this dissertation focuses on describing, documenting, and evaluating the origin story and system processes utilized to bring this initiative to implementation. The study describes the experiences of an initial group of in-service teachers seeking to complete a K-12 computer science teaching endorsement program in rural Wyoming by focusing on three overarching research questions: How does systems theory provide a model to deeply understanding Wyoming’s K-12 computer science education effort? What are the challenges of delivering multi-age computer science in the rural, mountain regions of the western U.S.? and What differences might exist between teachers’ perceived needs for computer science based professional development before and after their actual classroom teaching experience? Using a combination of standard quantitative and qualitative methods, the results suggest that the process could be analyzed using systems theory and that substantive differences exist in the suitability of various programming languages for in service teachers compared with the needs of traditional

关键词： Aging   Faces   Face recognition   Generative adversarial networks   Active appearance model   Computer science   Licenses   Face recognition   image generation   image database   face aging dataset   deep generative approach   generative adversarial network  

摘要： Many studies have been conducted in the field of face aging, from approaches that use pure image-processing algorithms, to those that use generative adversarial networks. In this study, we review a classic approach that uses a generative adversarial network. The structure, formulation, learning algorithm, challenges, advantages, and disadvantages of the algorithms contained in each proposed algorithm are discussed systematically. Generative Adversarial Networks are an approach that obtains the status of the art in the field of face aging by adding an aging module, paying special attention to the face part, and using an identity-preserving module to preserve identity. In this paper, we also discuss the database used for facial aging, along with its characteristics. The dataset used in the face aging process must have the following criteria: (1) a sufficiently large age group in the dataset, each age group must have a small range, (2) a balanced distribution of each age group, and (3) has enough number of face images.

ISBN: (纸本)9781477324196;9781477324202

摘要： 2023 Honorable Mention, Brazil Section Humanities BookPrize, Latin American Studies Association (LASA)This book explores visual portrayals of blackness in Brazi...

关键词： Clinical trials   Infectious diseases   Algorithms   Decision making   Markov analysis   Computer science   Operations research   Pharmaceutical sciences   Pharmacology  

摘要： The research in this dissertation is motivated by real applications of sequential decision making and reinforcement learning (RL). In numerous areas including clinical trials (Durand et al. 2018), finance (Shen et al. 2015), and recommendation systems (Zhou et al. 2017), we would like to take actions sequentially to gather more information and maximize the utility. It is crucial to have efficient and reliable statistical inference methods and learning algorithms to evaluate and optimize the policies which guide our decision making. With large and complicated datasets being widely available, it is of practical importance to develop meta-learning algorithms which can leverage information from multiple tasks, to design robust inference tools that provide performance guarantee on our policies in critical applications, and to propose RL frameworks that can handle multiple objectives and are applicable to vital areas such as public health. In Chapter 2, we study how to improve the learning efficiency in multi-armed bandits. Motivated by real applications, we introduce a novel metadata-based multi-task bandit problem, where the agent needs to solve a large number of related multi-armed bandit tasks and can leverage some task-specific features (i.e., metadata) to share knowledge across tasks. As a general framework, we propose to capture task relations through the lens of Bayesian hierarchical models, upon which a Thompson sampling algorithm is designed to efficiently learn task relations, share information, and minimize the cumulative regrets. Two concrete examples for Gaussian bandits and Bernoulli bandits are carefully analyzed. The Bayes regret for Gaussian bandits clearly demonstrates the benefits of information sharing with our algorithm. The proposed method is further supported by extensive experiments. In Chapter 3, we study how to provide efficient and robust estimation and inference procedures for off-policy evaluation in RL. Off-policy evaluation learns a target

关键词： Animal sciences   Artificial intelligence   Computer science   Electrical engineering   Engineering   Microbiology  

摘要： The overall objective of this dissertation was to investigate how feeding behavior and gut microbiome can be utilized to improve production in pig populations. A dataset obtained from Smithfield Premium Genetics contained purebred boars from three breeds (Duroc, Landrace, and Large White). Feed intake, feeder occupation time, body weight, and animal identification were recorded when a pig visited the Feed Intake Recording Equipment (FIRE) Feeder equipped in each pen during the feeding trial (70 to 169 days of age). The first study (Chapter 2) investigated the usefulness of electronic feeder data to predict the body weight of finishing pigs. We assessed the prediction performances of four models with different subsets of predictors (age, feed intake, and feeding behavior) within each of the four scenarios using three algorithms (LASSO regression, Random Forest, and Long Short-term Memory Network). Our results indicated that the contributions of feeding behavior and feed intake data to the prediction varied across scenarios, algorithms, and breeds considered. The remaining three studies (Chapters 3-5) included microbiome information (16S rRNA gene sequences) obtained by sampling rectal swabs from each animal at 73 ± 3, 123 ± 4, and 158 ± 4 days of age. The second study explored the interaction between gut microbiome and feeding behavior for each breed group at three growth stages. The proportion of phenotypic variance of feeding behavior measures attributed to variation in the gut microbiota composition (microbiability, m2 ) was small to moderate, ranging from 0.09 to 0.31 across breeds and growth stages. We identified 55 unique microbial taxa, mainly belonging to five phyla, with Firmicutes being the most abundant, significantly associated with different feeding behavior traits for the three breeds across the three sampling points along the growth. The last two studies (Chapters 4 and 5) included phenotypic records of growth and body composition, including body weigh

关键词： Capital punishment   Death penalty   Evangelium vitae   Human dignity   Catholic church  

摘要： Throughout the twentieth and twenty-first centuries, there has been a significant controversy within the Catholic Church regarding the moral permissibility of capital punishment. In light of Evangelium Vitae and recent magisterial documents, the death penalty has been condemned by Pope Francis as a deliberate attack on the dignity of the human person. This movement towards condemning the death penalty as illegitimate in all cases is in opposition to the teachings of Scripture, of the Church Fathers, and of Tradition that protect the state’s right to exercise just and proportionate punishments. The purpose of this essay is to demonstrate the progression of the doctrine on capital punishment from its Scriptural roots, as demonstrated by Aquinas, Augustine, and other prominent Catholic figures, to its current form by underlining its moral implications as defined by Pope John Paul II, Pope Francis, and the United States bishops. From this demonstration, I conclude that the teachings encouraged by Pope Francis must be read only in light of prudential judgement, not as a revision or development of doctrine. It is vital for the union of Catholics that the teachings and traditions of the Church are protected and accurately interpreted so that Catholics may properly advocate for the sanctity of human dignity and create a culture of life under the Church’s guidance.

关键词： Artificial intelligence   Computer science  

关键词： Training   Task analysis   Semantics   Shape   Representation learning   Generative adversarial networks   Computer science   Person re-identification   cloth-changing   data augmentation  

摘要： This paper studies the challenging person re-identification (Re-ID) task under the cloth-changing scenario, where the same identity (ID) suffers from uncertain cloth changes. To learn cloth- and ID-invariant features, it is crucial to collect abundant training data with varying clothes, which is difficult in practice. To alleviate the reliance on rich data collection, we reinforce the feature learning process by designing powerful complementary data augmentation strategies, including positive and negative data augmentation. Specifically, the positive augmentation fulfills the ID space by randomly patching the person images with different clothes, simulating rich appearance to enhance the robustness against clothes variations. For negative augmentation, its basic idea is to randomly generate out-of-distribution synthetic samples by combining various appearance and posture factors from real samples. The designed strategies seamlessly reinforce the feature learning without additional information introduction. Extensive experiments conducted on both cloth-changing and -unchanging tasks demonstrate the superiority of our proposed method, consistently improving the accuracy over various baselines.

关键词： Systematics   Bibliographies   Data integrity   Data mining   Computer science   STEM   Market research   Systematic literature review   data imputation   data mining   missingness   data preprocessing   data quality  

摘要： Missing data are a universal data quality problem in many domains, leading to misleading analysis and inaccurate decisions. Much research has been done to investigate the different mechanisms of missing data and the proper techniques in handling various data types. In the last decade, machine learning has been utilized to replace conventional methods to address the problem of missing values more efficiently. By studying and analyzing recently proposed methods using machine learning approaches, vital adoptions in accuracy, performance, and time consumed can be highlighted. This study aimed to help data analysts and researchers address the limitations of machine learning imputation methods by conducting a systematic literature review to provide a comprehensive overview of using such methods to impute missing values. Novel proposed machine learning approaches used for data imputation are analyzed and summarized to assist researchers in selecting a proper machine learning method based on several factors and settings. The review was performed on research studies published between 2016 and 2021 on adopting machine learning to impute missing values, focusing on their strengths and limitations. A total of 684 research articles from various scientific databases were analyzed using search engines, and 94 of them were selected as primary studies. Finally, several recommendations were given to guide future researchers in applying machine learning to impute missing values.