教学课程资源库 更多>>

关键词： Feature extraction   Convolution   Quantization (signal)   Training   Task analysis   Computer science   Hyperspectral imaging   Binary neural network (BNN)   hyperspectral image (HSI) classification   progressive quantization  

摘要： Convolutional neural network (CNN) has shown its powerful ability for hyperspectral image (HSI) classification, which however, is difficult to deploy on resource-limited or low-latency platforms due to its parameter and computation redundancy. Though binary neural network (BNN) has attracted attention for its extreme compressing and speeding up ability by binarizing both weights and activations, it has rarely been explored for HSI classification. In this study, we elaborately design a BNN with good performance for HSI classification task. Specifically, an adaptive gradient scale module is proposed to flexibly modify the gradient during training stage to better optimize the BNN and does not add any extra computation for inference. Furthermore, a curriculum learning-based progressive binarization strategy is utilized to improve the performance. Compared with the existing BNN works, our method can increase the HSI classification accuracy by a large margin while maintaining the compressing ratio. Abundant experiments on three datasets demonstrate the effectiveness of the proposed method.

关键词： Career   Computer   Identity   Middle   Science  

摘要： The purpose of this study was to understand whether students who took computer science courses in a prescribed sequence during middle school developed a strong enough computing identity to show an interest in continuing to take computing courses in high school and possibly pursue a career in software development. This study was quantitative and non-experimental. The participants consisted of 184 sixth through eighth grade students, across 15 middle schools enrolled in one of five computer science courses in a large urban district in the southeast region of the United States. The instrument used to analyze a student’s overall computing identity was a survey form that consisted of 11 statements, of which nine were slightly modified from the model research by Mahadeo et al. (2020). Two questions were added to investigate a student’s aspirations to take more software development in high school and their intention of pursuing a career in software development. All questions were answered on a five-point Likert-type scale. Six research questions were constructed for this study to compare computing identity development regarding courses, pathways, race, Title I status, and interest in coding beyond middle school. This study used descriptive statistics, F-test, and ANOVA to capture a broad understanding of the development of computing identity in middle school students who were taking computer science courses in a sequenced pathway. There were three findings: modification of the statements did not impact the overall structure of the tool, computer science pathways were not implemented with fidelity, and there was a strong likelihood students with a high computing identity would also have a high interest in taking more courses in high school and pursuing a career in software development.

关键词： Encryption   Cryptography   Noise measurement   Error correction   Computer science   Visualization   Read only memory   Cryptcoding   error-correcting codes   burst channel   image   quasigroups  

摘要： The concept of cryptcoding arises from the need to obtain secure and accurate transmission. This has led to an intensive development of coding theory and cryptography as scientific fields dealing with these problems. To ensure efficient and secure data transmission at the same time, the concept of cryptcoding is being developed, in which the coding and encryption processes are merged into one process. Cryptcodes provide correction of a certain number of errors in the transmitted message and data confidentiality, using only one algorithm. In this paper, we consider cryptcodes based on quasigroups, proposed elsewhere. Also, Burst-Cut-Decoding, Burst-4-Sets-Cut-Decoding, FastB-Cut-Decoding and FastB-4-Sets-Cut-Decoding algorithms for transmission of messages through burst channels have been defined and investigated elsewhere. Here, we investigate performances of these algorithms for transmission of images through burst channels. We made experiments for different channel parameters and compared the results obtained with different decoding algorithms of these cryptcodes. In all experiments, we considered the bit-error probability and the differences between the transmitted and decoded images. From the results presented in this paper we can conclude that Fast algorithms improve the performances of these cryptcodes for transmission of images over a burst channel. The best results are obtained by FastB-4-Sets-Cut-Decoding algorithm. In addition, to enhance the quality of the decoded images, we examined the application of a filter for visual correction of unsuccessfully decoded pixels using the surrounding pixels. The considered cryptcodes and presented results can be useful for application in satellite digital video broadcasting (DVB-S) coding and encryption schemes.

关键词： Aging   Cognitive load theory   Computer Science   Lifelong learning   Lifespan theory   Recommender system  

摘要： Education must be a lifelong pursuit. The expansion of knowledge propagates significant cultural change within the average lifespan, and knowledge currency is brief. Extended education may slow the cognitive decline of healthy aging and dementia. However, there are barriers to lifelong learning for adults over 65 years old. The purpose of this quantitative quasi-experimental study was to examine the impact of a course recommender system, introduced as an alternative to the course catalog, on perceived cognitive load in lifelong learners over 65 years old in a university community-based lifelong learning program in the Southeastern United States. The study involved a stratified sample of 56 participants randomly assigned into experimental and control groups. Both groups used the treatment software with a different user experience (UX). The experimental UX was a course recommender system. The control UX was an electronic course catalog. Sweller's (1988) cognitive load theory posits that working memory has a finite capacity, and extraneous cognitive load consumes limited resources. Descriptive statistics suggest that a study repeat should further substantiate the primary hypothesis, as perceived mental effort was lower for the experimental group. The single usability metric (SUM; Sauro & Kindlund, 2005) indicated that the system was usable, and the i-bar statistic (Hoffman et al., 2010) indicated that the system was learnable. Social contact was again shown to be a key factor in seniors’ educational choices (Dattilo et al., 2012). Future research should span two semesters, incorporating a prediction algorithm into the system for testing during the second semester.

关键词： Women students   Student motivations   College students   Computer science classes   Goal orientation theory  

摘要： The purpose of this qualitative descriptive study is to explore how college students from the United States describe the motivational factors that influence them to seek a degree in computer science. There is a gap in the research that shows a need to look at the motivations of women from their first experiences in computer science to begin and stay on a path to graduate with a degree. This study is directly connected to goal orientation theory because the research questions being asked will try to answer the four types of goal orientations. The research questions identified for this study had participants answer the semi-structured interview questions looking at different motivations and the unique experiences in their journey towards pursuing and completing a Computer Science degree. Purposeful sampling was used to recruit 15 women computer science students in the junior or senior of college. Data was collected by ten of the participants being interviewed with the other five joining together for a focus group. Through thematic analysis, five themes were identified: wanting a good job; wanting to earn good grades, fitting in with peers, and their love of computers. According to this study, there are specific internal factors, external rewards or practices and the aspects associated with computer science classes and assignments that motivate women college students.

关键词： Image color analysis   Indexes   Green products   Optimization   Computer science   Finite element analysis   Reflectivity   Color selection   color space   Hadamard product   Heaviside step function  

摘要： Numerous research endeavors have employed color selection procedures for a vast array of purposes. Detecting defects in fabrics, calculating the Microbial Community Color Index, analyzing digital color to facilitate fashion design processes, applying color to artworks, calculating canopy cover, and achieving other objectives have been the subject of research. This procedure requires intricate steps, calculations, and a lengthy computational time. In this study, a novel strategy for optimizing the color selection process using the Hadamard product technique is presented. The HSV color space is optimized by selectively selecting the desired colors and establishing threshold limits for each hue, saturation, and value component. The optimization results demonstrate that the desired colors are perfectly distinguished from other colors. Additionally, the proposed method employs a straightforward, step-by-step procedure that does not require feature extraction. In comparison to previous research, a remarkable increase in computational speed of 1,078.82 times faster has been observed. This improvement is achieved by multiplying each element of the HSV matrix resulting from color selection as opposed to the HSV matrix without selection. The faster computational speed observed in this study during the color selection process has the potential to be used in unmanned aerial vehicles to select green plants or ripe *** study's findings are applicable not only to plant images but also to all cases requiring color selection under visible light conditions.

关键词： Computer Science   Mathematics \u0026 Statistics  

ISBN: (数字)9788770221597 ISBN: (纸本)9788770221603

摘要： This book explores the fundamentals and applications of Blockchain technology. Readers will learn about the decentralized peer-to-peer network, distributed ledger, and the trust model that defines Blockchain technology. They will also be introduced to the basic components of Blockchain (transaction, block, block header, and the chain), its operations (hashing, verification, validation, and consensus model), underlying algorithms, and essentials of trust (hard fork and soft fork). Private and public Blockchain networks similar to Bitcoin and Ethereum will be introduced, as will concepts of Smart Contracts, Proof of Work and Proof of Stack. Blockchain is an emerging technology that can radically improve transaction security at banking, supply chain, and other transaction networks. It’s estimated that Blockchain will generate $3.1 trillion in new business value by 2030. Essentially, it provides the basis for a dynamic distributed ledger that can be applied to save time when recording transactions between parties, remove costs associated with intermediaries, and reduce risks of fraud and tampering.

关键词： Perturbation methods   Artificial neural networks   Laser beams   Face recognition   Training   Robustness   Computer science   Adversarial detection   deep learning attacks   label sequences discrepancy   activation spaces  

摘要： Deep neural network (DNN) models have been widely used in many tasks due to their superior performance. However, DNN models are usually vulnerable to adversarial example attacks, which limits their applications in many safety-critic scenarios. How to effectively detect adversarial examples to enhance the robustness of DNN models has attracted much attention in recent years. Most adversarial example detection methods require modifying or retraining the model, which is impractical and reduces the classification accuracy of normal examples. In this paper, we propose an adversarial example detection approach that does not require modification of the DNN models and meanwhile retains the classification accuracy of normal examples. The key observation is that when we transform the input example with some operations (e.g., masking a pixel with a reference value), feed the transformed example to the target model, and use the output of the intermediate layers to predict the label of the example, the generated label sequences of adversarial examples will be extremely discrepant but the label sequences of normal examples keep nearly unchanged. Motivated by this observation, we design an approach to detect adversarial examples based on the label sequence discrepancy (LSD) of the given examples. The experimental results against five mainstream adversarial attacks on three benchmark datasets demonstrate that LSD outperforms the state-of-the-art solutions in the detection rate of adversarial examples. Moreover, LSD performs well at various confidence levels and exhibits good generalizability between different attacks.

关键词： Hidden Markov models   Acoustics   Mathematical models   Feature extraction   Mel frequency cepstral coefficient   Computer science   Computational modeling   Pronunciation scoring   goodness of pronunciation   hidden Markov model-deep neural network   dynamic time warping   kaldi   speechocean762  

摘要： The current pronunciation scoring based on Goodness of Pronunciation (GOP) uses posterior probabilities of the Acoustic Models. Such algorithms suffer from generalization since they are utilized to determine a score metric for each phoneme rather than on the completeness or comparison with the ideal utterance of the words. This paper proposes a novel method to overcome such limitations by using combined scores of prosodic, fluency, completeness, and accuracy. This is achieved using context-aware GOP in conjugation with dynamic time warping (DTW) matching of the pitch contours of a weighted average of the context tokens found in the audio file that is rich in mispronounced phonemes. The proposed work gives flexibility in tuning the results according to different speech aspects based on a single hyperparameter. The results achieved are encouraging and have been validated on the speechocean762 dataset, where Automatic Speech Recognition (ASR) model has been trained on the Librispeech dataset. The resultant mean error of the proposed approach is 3.38% and the value of the correlation coefficient achieved is 0.652.

关键词： 英语学习活动观   阅读教学   核心素养  

摘要： 《义务教育课程标准（2022年版）》指出英语教学要围绕核心素养开展，践行英语学习活动观是培育核心素养的重要途径。本文以鲁教版初中英语七年级下册Unit 2“I'm going to study computer science.” Section B阅读教学为课例，探析如何在初中英语阅读教学中践行英语学习活动观，培养学生的英语核心素养，实现学科育人目标。