• Asia Pacific Journal of Corpus Research
• /
• v.5 no.1
• /
• pp.23-36
• /
• 2024

This research centers on the Taiwan Eye-Movement Corpus of Spanish (TECS), a specially created corpus comprising eye-tracking data from Chinese-speaking learners of Spanish as a third language in Taiwan. Its primary purpose is to explore the broad utility of TECS in understanding language learning processes, particularly the initial stages of language learning. Constructing this corpus involves gathering data on eye-tracking, reading comprehension, and language proficiency to develop a machine-learning model that predicts learner behaviors, and subsequently undergoes a predictability test for validation. The focus is on examining attention in input processing and their relationship to language learning outcomes. The TECS eye-tracking data consists of indicators derived from eye movement recordings while reading Spanish sentences with temporal references. These indicators are obtained from eye movement experiments focusing on tense verbal inflections and temporal adverbs. Chinese expresses tense using aspect markers, lexical references, and contextual cues, differing significantly from inflectional languages like Spanish. Chinese-speaking learners of Spanish face particular challenges in learning verbal morphology and tenses. The data from eye movement experiments were structured into feature vectors, with learner behaviors serving as class labels. After categorizing the collected data, we used two types of machine learning methods for classification and regression: Random Forests and the k-nearest neighbors algorithm (KNN). By leveraging these algorithms, we predicted learner behaviors and conducted performance evaluations to enhance our understanding of the nexus between learner behaviors and language learning process. Future research may further enrich TECS by gathering data from subsequent eye-movement experiments, specifically targeting various Spanish tenses and temporal lexical references during text reading. These endeavors promise to broaden and refine the corpus, advancing our understanding of language processing.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.3
• /
• pp.127-133
• /
• 2019

Most studies of software fault prediction have been about supervised learning models that use only labeled training data. Although supervised learning usually shows high prediction performance, most development groups do not have sufficient labeled data. Unsupervised learning models that use only unlabeled data for training are difficult to build and show poor performance. Semi-supervised learning models that use both labeled data and unlabeled data can solve these problems. Self-training technique requires the fewest assumptions and constraints among semi-supervised techniques. In this paper, we implemented several models using self-training algorithms and evaluated them using Accuracy and AUC. As a result, YATSI showed the best performance.

• The Journal of the Convergence on Culture Technology
• /
• v.7 no.1
• /
• pp.615-619
• /
• 2021

This study focuses on memory systems that are optimized to increase performance and energy efficiency in many embedded systems such as IoT, cloud computing, and edge computing, and proposes a performance improvement technique. The proposed technique improves memory system performance based on machine learning algorithms that are widely used in many applications. The machine learning technique can be used for various applications through supervised learning, and can be applied to a data classification task used in improving memory system performance. Data classification based on highly accurate machine learning techniques enables data to be appropriately arranged according to data usage patterns, thereby improving overall system performance.

• Journal of Korea Multimedia Society
• /
• v.24 no.3
• /
• pp.382-395
• /
• 2021

In this study, character recognition using deep learning is performed among the various defects in the PCB, the purpose of which is to check whether the printed characters are printed correctly on top of components, or the incorrect parts are attached. Generally, character recognition may be perceived as not a difficult problem when considering MNIST, but the printed letters on the PCB component data are difficult to collect, and have very high redundancy. So if a deep learning model is trained with original data without any preprocessing, it can lead to over fitting problems. Therefore, this study aims to reduce the redundancy to the smallest dataset that can represent large amounts of data collected in limited production sites, and to create datasets through data enhancement to train a flexible deep learning model can be used in various production sites. Moreover, ResNet model verifies to determine which combination of datasets is the most effective. This study discusses how to reduce and augment data that is constantly occurring in real PCB production lines, and discusses how to select coresets to learn and apply deep learning models in real sites.

• International Journal of Computer Science & Network Security
• /
• v.22 no.4
• /
• pp.119-130
• /
• 2022

Crime is a common social problem that affects the quality of life. As the number of crimes increases, it is necessary to build a model to predict the number of crimes that may occur in a given period, identify the characteristics of a person who may commit a particular crime, and identify places where a particular crime may occur. Data privacy is the main challenge that organizations face when building this type of predictive models. Federated learning (FL) is a promising approach that overcomes data security and privacy challenges, as it enables organizations to build a machine learning model based on distributed datasets without sharing raw data or violating data privacy. In this paper, a federated long short- term memory (LSTM) model is proposed and compared with a traditional LSTM model. Proposed model is developed using TensorFlow Federated (TFF) and the Keras API to predict the number of crimes. The proposed model is applied on the Boston crime dataset. The proposed model's parameters are fine tuned to obtain minimum loss and maximum accuracy. The proposed federated LSTM model is compared with the traditional LSTM model and found that the federated LSTM model achieved lower loss, better accuracy, and higher training time than the traditional LSTM model.

• Journal of Information Processing Systems
• /
• v.19 no.5
• /
• pp.673-687
• /
• 2023

Convolutional neural network-based deep learning technology is the most commonly used in image identification, but it requires large-scale data for training. Therefore, application in specific fields in which data acquisition is limited, such as in the military, may be challenging. In particular, the identification of ground weapon systems is a very important mission, and high identification accuracy is required. Accordingly, various studies have been conducted to achieve high performance using small-scale data. Among them, the ensemble method, which achieves excellent performance through the prediction average of the pre-trained models, is the most representative method; however, it requires considerable time and effort to find the optimal combination of ensemble models. In addition, there is a performance limitation in the prediction results obtained by using an ensemble method. Furthermore, it is difficult to obtain the ensemble effect using models with imbalanced classification accuracies. In this paper, we propose a transfer learning-based feature fusion technique for heterogeneous models that extracts and fuses features of pre-trained heterogeneous models and finally, fine-tunes hyperparameters of the fully connected layer to improve the classification accuracy. The experimental results of this study indicate that it is possible to overcome the limitations of the existing ensemble methods by improving the classification accuracy through feature fusion between heterogeneous models based on transfer learning.

• Journal of KIBIM
• /
• v.13 no.3
• /
• pp.12-20
• /
• 2023

Deep learning-based anomaly detection technology is used in various fields such as computer vision, speech recognition, and natural language processing. In particular, this technology is applied in various fields such as monitoring manufacturing equipment abnormalities, detecting financial fraud, detecting network hacking, and detecting anomalies in medical images. However, in the field of construction and architecture, research on deep learning-based data anomaly detection technology is difficult due to the lack of digitization of domain knowledge due to late digital conversion, lack of learning data, and difficulties in collecting and processing field data in real time. This study acquires necessary data through IoT (Internet of Things) from the viewpoint of monitoring for environmental management of architectural spaces, converts them into a database, learns deep learning, and then supports anomaly patterns using AI (Artificial Infelligence) deep learning-based anomaly detection. We propose an implementation process. The results of this study suggest an effective environmental anomaly pattern detection solution architecture for environmental management of architectural spaces, proving its feasibility. The proposed method enables quick response through real-time data processing and analysis collected from IoT. In order to confirm the effectiveness of the proposed method, performance analysis is performed through prototype implementation to derive the results.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1408-1414
• /
• 2019

Deep artificial neural network with transfer learning is applied to compressed sensing cardiovascular MRI. Transfer learning is a method that utilizes structure, filter kernels, and weights of the network used in prior learning for current learning or application. The transfer learning is useful in accelerating learning speed, and in generalization of the neural network when learning data is limited. From a cardiac MRI experiment, with 8 healthy volunteers, the neural network with transfer learning was able to reduce learning time by a factor of more than five compared to that with standalone learning. Using test data set, reconstructed images with transfer learning showed lower normalized mean square error and better image quality compared to those without transfer learning.

• Educational Technology International
• /
• v.20 no.2
• /
• pp.195-221
• /
• 2019

This study explored middle and high school students' learning experiences using e-textbooks in online learning courses. Data were collected from in-depth interviews. The interviewees for this study were 19 students who enrolled voluntarily in an online mathematics and science inquiry program, actively participated in the online learning. The students generally have high academic achievement and motivation for learning in science and mathematics. Data were analyzed based on a grounded theory approach. As a result, the characteristics of the online learning environment using e-textbooks were conceptualized via three different categories including temporal, spatial, and technical. Such characteristics of the learning environment were able to provoke self-directed learning, extended learning, interactive learning, in-depth learning, improved ICT literacy, and formation of positive emotions and learning habits. Most of the learners showed positive feedback towards the use of e-textbooks, while some mentioned the technical limitations compared to conventional paper-based learning. This study suggested that e-textbooks are likely to induce positive experiences for learners in the context of online learning, so it is necessary to design contents that utilize various functions and advantages of electronic teaching materials in order to use e-textbooks effectively.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.142-145
• /
• 2003

Many control applications using Neural Network need a priori information about the objective system. But it is impossible to get exact information about the objective system in real world. To solve this problem, several control methods were proposed. Reinforcement learning control using neural network is one of them. Basically reinforcement learning control doesn't need a priori information of objective system. This method uses reinforcement signal from interaction of objective system and environment and observable states of objective system as input data. But many methods take too much time to apply to real-world. So we focus on faster learning to apply reinforcement learning control to real-world. Two data types are used for reinforcement learning. One is reinforcement signal data. It has only two fixed scalar values that are assigned for each success and fail state. The other is observable state data. There are infinitive states in real-world system. So the number of observable state data is also infinitive. This requires too much learning time for applying to real-world. So we try to reduce the number of observable states by classification of states with Self-Organizing Map. We also use neural dynamic programming for controller design. An inverted pendulum on the cart system is simulated. Failure signal is used for reinforcement signal. The failure signal occurs when the pendulum angle or cart position deviate from the defined control range. The control objective is to maintain the balanced pole and centered cart. And four states that is, position and velocity of cart, angle and angular velocity of pole are used for state signal. Learning controller is composed of serial connection of Self-Organizing Map and two Multi-layer Feed-Forward Neural Networks.