• 아동학회지
• /
• 제4권
• /
• pp.37-51
• /
• 1983

The purpose of this study was the investigation of theoretical and empirical relationships between Piagetian and psychometric assessments of intelligence. Specifically, the factor structure of Piagetian-type scales, the relationship between Piagetian scales and psychometric intelligence tests, and differences in the factor structure of Piagetian and psychometric assessments of intelligence were studied. The subjects of this stuby were 70 children (35 boys and 35 girls) in the 1st grade of an elementary school in Seoul The Piagetian-type scales and the K-WISC were administered individually, and the General Intelligence Test was administered to groups of children. Statistical analysis of the obtained data consisted of the SPSS Computer program including factor analysis and Pearson's product moment correlation coefficient. The Piagetian-type scales were found to consist of three factors, which accounted for 55 percent of the total common-factor variance. Factor-I was a factor indicating "conservation". Factor-II was a factor indicating "moral judgements". Factor-III was a factor indicating "classification and identity". Correlations between subtests of psychometric tests and Piagetian scales were relatively low or moderate. Relations between IQs assessed by the psychometric tests and Piagetian scales were also relativeyly low or moderate. Eight factors were extracted from the joint factor analysis of psychometric intelligence tests and Piagetian scales, and they accounted for 67 percent of the total common-factor variance. Factors-I, II, III, and V consisted of subtests of psychometric assessments, and Factors-IV, VI, VII and VIII were composed of Piagetian scales. Factor-I was a factor for "reasoning ability based upon language". Factor-II was a factor for "performance ability". Factor-III was a factor for "grouping ability". Factor-IV was a factor for "conservation". Factor-V was a factor indicating "symbol and language usage ability". Factor- VI was a factor indicating "moral judgments". Factor-VII was a factor indicating "length consevation". Factor-VIII was a factor indicating "classification and identity".

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권6호
• /
• pp.2806-2825
• /
• 2018

Unsupervised learning has shown good performance on image, video and audio classification tasks, and much progress has been made so far. It studies how systems can learn to represent particular input patterns in a way that reflects the statistical structure of the overall collection of input patterns. Many promising deep learning systems are commonly trained by the greedy layerwise unsupervised learning manner. The performance of these deep learning architectures benefits from the unsupervised learning ability to disentangling the abstractions and picking out the useful features. However, the existing unsupervised learning algorithms are often difficult to train partly because of the requirement of extensive hyperparameters. The tuning of these hyperparameters is a laborious task that requires expert knowledge, rules of thumb or extensive search. In this paper, we propose a simple and effective unsupervised feature learning algorithm for image classification, which exploits an explicit optimizing way for population and lifetime sparsity. Firstly, a sparse target matrix is built by the competitive rules. Then, the sparse features are optimized by means of minimizing the Euclidean norm ($L_2$) error between the sparse target and the competitive layer outputs. Finally, a classifier is trained using the obtained sparse features. Experimental results show that the proposed method achieves good performance for image classification, and provides discriminative features that generalize well.

• 한국멀티미디어학회논문지
• /
• 제19권7호
• /
• pp.1197-1207
• /
• 2016

The purpose of this study was to develop an evidence-based Nursing Process System on Nursing Diagnosis, Nursing Outcomes, and Nursing Interventions Classification targeting nurse students. We use standard classification-focused research data on the basis of Nursing Diagnosis Classification established by NANDA (North American Nursing Diagnosis Association), NOC (Nursing Outcomes Classification) and NIC (Nursing Interventions Classification) mainly developed by Iowa Sate University. The existing research methods are difficult to be applied the consistent nursing process, since such methods need to repeatedly enter the same nursing process without systematic guidelines. But, this study was coded data of standardized nursing process in accordance with the 10 clinical condition in order to implement the nursing process macro, and developed a system that reflects the needs of nursing educators. Therefore, nurse students can improve clinical decision-making ability, and naturally learn the nursing process through a system developed.

• 대한물리의학회지
• /
• 제14권2호
• /
• pp.97-105
• /
• 2019

PURPOSE: This study examined the factors related to school-based activity performance in school-aged children with spastic cerebral palsy (CP). METHODS: The Gross Motor Function Systems (GMFCS), Manual Ability Classification System (MACS), Communication Function Classification System (CFCS) as functional classifications, and the physical activity performance of the School Function Assessment (SFA) were measured in 79 children with spastic CP to assess the student's performance of specific school-related functional activities. RESULTS: All the function classification systems were correlated significantly with the physical activity performance of the SFA ($r_s=-.47$ to -.80) (p<.05). The MACS (${\beta}=-.59$), GMFCS (${\beta}=-.23$), CFCS (${\beta}=-.21$), and age (${\beta}=-.15$) in order were predictors of the physical activity performance of the SFA (84.8%)(p<.05). CONCLUSION: These functional classification systems can be used to predict the school-based activity performance in school-aged children with CP. In addition, they can contribute to the selection of areas for intensive interventions to improve the school-based activity performance.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2019년도 춘계학술대회
• /
• pp.442-444
• /
• 2019

직물결함 분류는 원단 품질관리에 있어 중요한 문제이다. 하지만, 다양한 결함의 종류를 영상으로 식별하기 어렵기 때문에 자동화가 어렵다. 따라서 직물결함 분류는 대부분 사람에게 의존하고 있다. 본 논문에서는, 이를 해결하기 위해 직물결함 분류 문제에 CNN을 적용한다. 또한 CNN의 학습을 보다 쉽게 하기 위하여, 사람이 영상에 결함 영역을 표시하는 방법을 제안한다. 본 논문에서는 제안방법과 원본영상에 대한 비교실험을 수행하여, 제안방법이 학습에 효과가 있다는 것을 확인하였다.

• Journal of Information Processing Systems
• /
• 제19권6호
• /
• pp.830-841
• /
• 2023

Tokenization is the process of segmenting the input text into smaller units of text, and it is a preprocessing task that is mainly performed to improve the efficiency of the machine learning process. Various tokenization methods have been proposed for application in the field of natural language processing, but studies have primarily focused on efficiently segmenting text. Few studies have been conducted on the Korean language to explore what tokenization methods are suitable for document classification task. In this paper, an exploratory study was performed to find the most suitable tokenization method to improve the performance of a representative spatio-temporal document classifier in Korean. For the experiment, a convolutional neural network model was used, and for the final performance comparison, tasks were selected for document classification where performance largely depends on the tokenization method. As a tokenization method for comparative experiments, commonly used Jamo, Character, and Word units were adopted. As a result of the experiment, it was confirmed that the tokenization of word units showed excellent performance in the case of representative spatio-temporal document classification task where the semantic embedding ability of the token itself is important.

• 대한산업공학회지
• /
• 제39권5호
• /
• pp.393-402
• /
• 2013

The Mahalanobis-Taguchi System (MTS) is a diagnostic and predictive method for multivariate data. In the MTS, the Mahalanobis space (MS) of reference group is obtained using the standardized variables of normal data. The Mahalanobis space can be used for multi-class classification. Once this MS is established, the useful set of variables is identified to assist in the model analysis or diagnosis using orthogonal arrays and signal-to-noise ratios. And other several techniques have already been used for classification, such as linear discriminant analysis and logistic regression, decision trees, neural networks, etc. The goal of this case study is to compare the ability of the Mahalanobis-Taguchi System and logistic regression using a data set.

• Journal of Multimedia Information System
• /
• 제3권3호
• /
• pp.53-61
• /
• 2016

This research proposes the index optimization as a classification task and application of the graph based KNN. We need the index optimization as an important task for maximizing the information retrieval performance. And we try to solve the problems in encoding words into numerical vectors, such as huge dimensionality and sparse distribution, by encoding them into graphs as the alternative representations to numerical vectors. In this research, the index optimization is viewed as a classification task, the similarity measure between graphs is defined, and the KNN is modified into the graph based version based on the similarity measure, and it is applied to the index optimization task. As the benefits from this research, by modifying the KNN so, we expect the improvement of classification performance, more graphical representations of words which is inherent in graphs, the ability to trace more easily results from classifying words. In this research, we will validate empirically the proposed version in optimizing index on the two text collections: NewsPage.com and 20NewsGroups.

• 전기학회논문지
• /
• 제64권1호
• /
• pp.113-120
• /
• 2015

In this paper, we propose a feature extraction method using the stacked autoencoders which consist of restricted Boltzmann machines. The stacked autoencoders is a sort of deep networks. Restricted Boltzmann machines (RBMs) are probabilistic graphical models that can be interpreted as stochastic neural networks. In terms of pattern classification problem, the feature extraction is a key issue. We use the stacked autoencoders networks to extract new features which have a good influence on the improvement of the classification performance. After feature extraction, fuzzy k-nearest neighbors algorithm is used for a classifier which classifies the new extracted data set. To evaluate the classification ability of the proposed pattern classifier, we make some experiments with several machine learning data sets.

• 대한산업공학회지
• /
• 제37권4호
• /
• pp.400-407
• /
• 2011

Modeling and analysis of high-dimensional spectral data provide an opportunity to uncover inherent patterns in various information-rich data. Orthogonal signal correction (OSC) a preprocessing technique has been widely used to remove unwanted variations of spectral data that do not contribute to prediction or classification. In the present study we propose a novel OSC algorithm called adjusted direct OSC to improve visualization and the ability of classification. Experimental results with real mass spectral data from condom lubricants demonstrate the effectiveness of the proposed approach.