• The Mathematical Education
• /
• v.63 no.2
• /
• pp.165-186
• /
• 2024

Despite the growing attention on artificial intelligence-based automated scoring technology as a support method for the introduction of descriptive items in school environments and large-scale assessments, there is a noticeable lack of foundational research in mathematics compared to other subjects. This study developed an automated scoring model for two descriptive items in first-year middle school mathematics using the Random Forest algorithm, evaluated its performance, and explored ways to enhance this performance. The accuracy of the final models for the two items was found to be between 0.95 to 1.00 and 0.73 to 0.89, respectively, which is relatively high compared to automated scoring models in other subjects. We discovered that the strategic selection of the number of evaluation categories, taking into account the amount of data, is crucial for the effective development and performance of automated scoring models. Additionally, text preprocessing by mathematics education experts proved effective in improving both the performance and interpretability of the automated scoring model. Selecting a vectorization method that matches the characteristics of the items and data was identified as one way to enhance model performance. Furthermore, we confirmed that oversampling is a useful method to supplement performance in situations where practical limitations hinder balanced data collection. To enhance educational utility, further research is needed on how to utilize feature importance derived from the Random Forest-based automated scoring model to generate useful information for teaching and learning, such as feedback. This study is significant as foundational research in the field of mathematics descriptive automatic scoring, and there is a need for various subsequent studies through close collaboration between AI experts and math education experts.

• Asian Pacific Journal of Cancer Prevention
• /
• v.13 no.1
• /
• pp.147-155
• /
• 2012

Aboriginal and Torres Strait Islander people comprise about 2.5% of the Australian population. Cancer registry data indicate that their breast cancer survivals are lower than for other women but the completeness and accuracy of Indigenous descriptors on registries are uncertain. We followed women receiving mammography screening in BreastScreen to determine differences in screening experiences and survivals from breast cancer by Aboriginal and Torres Strait Islander status, as recorded by BreastScreen. This status is self-reported and used in BreastScreen accreditation, and is considered to be more accurate. The study included breast cancers diagnosed during the period of screening and after leaving the screening program. Design: Least square regression models were used to compare screening experiences and outcomes adjusted for age, geographic remoteness, socio-economic disadvantage, screening period and round during 1996-2005. Survival of breast cancer patients from all causes and from breast cancer specifically was compared for the 1991-2006 diagnostic period using linked cancer-registry data. Cox proportional hazards regression was used to adjust for socio-demographic differences, screening period, and where available, tumour size, nodal status and proximity of diagnosis to time of screen. Results: After adjustment for socio-demographic differences and screening period, Aboriginal and Torres Strait Islander women participated less frequently than other women in screening and re-screening although this difference appeared to be diminishing; were less likely to attend post-screening assessment within the recommended 28 days if recalled for assessment; had an elevated ductal carcinoma in situ but not invasive cancer detection rate; had larger breast cancers; and were more likely than other women to be treated by mastectomy than complete local excision. Linked cancer registry data indicated that five-year year survivals of breast cancer cases from all causes of death were 81% for Aboriginal and Torres Strait Islander women, compared with 90% for other women, and that the former had larger breast cancers that were more likely to have nodal spread at diagnosis. After adjusting for socio-demographic factors, tumour size, nodal spread and time from last screen to diagnosis, Aboriginal and Torres Strait Islander women had approximately twice the risk of death from breast cancer as other women. Conclusions: Aboriginal and Torres Strait Islander women have less favourable screening experiences and those diagnosed with breast cancer (either during the screening period or after leaving the screening program) have lower survivals that persist after adjustment for socio-demographic differences, tumour size and nodal status.

• Journal of Chest Surgery
• /
• v.30 no.11
• /
• pp.1092-1096
• /
• 1997

Thromboelastography(TEG) enables a global assessment of hemostatic function to be made from a single blood sample, documenting the interaction of platelets with protein coagulation cascade from the time of the initial platelet-fibrin interaction, through platelet aggregation, clot strengthening and fibrin cross linking to eventual clot Iysis. Thirty-five patients(mean age 34$\pm$ 12) undergoing open heart surgery from April 1st, 1996 to August 31th, 1996 were investigated at preoperatively and immediate, one hour, and 24 hours after cessation of cardiopulmonary bypass using TEG. Comparisons were made between classic hematological indices and TEG data. There were statistically significant correlation between maximal amplitude(MA) and platelet count before CPB, activating clotting time(ACT) and TEG date(R time, K time and a angle) at 24-hour after CPB. The data on the predictive accuracy for postoperative bleeding at 24-hour after CPB, the TEG was significantly better than ACT(57%) or the coagulation profiles(43%) as a predictor of postoperative bleeding, with an accuracy rate of 100% (P=0.0043). In conclusion, TEG seems to be easy to use, clinically accurate, cost effective and provides data which can effectively manage a patient's hemostasis.

• Korean Journal of Remote Sensing
• /
• v.36 no.5_4
• /
• pp.1179-1194
• /
• 2020

This study investigates the feasibility of three algorithms, K-Nearest Neighbors (K-NN), Random Forest (RF) and Neural Network (NN), for estimating the air temperature of an unobserved area where the weather station is not installed. The satellite image were obtained from Landsat-8 and MODIS Aqua/Terra acquired in 2019, and the meteorological ground weather data were from AWS/ASOS data of Korea Meteorological Administration and Korea Forest Service. In addition, in order to improve the estimation accuracy, a digital surface model, solar radiation, aspect and slope were used. The accuracy assessment of machine learning methods was performed by calculating the statistics of R² (determination coefficient) and Root Mean Square Error (RMSE) through 10-fold cross-validation and the estimated values were compared for each target area. As a result, the neural network algorithm showed the most stable result among the three algorithms with R² = 0.805 and RMSE = 0.508. The neural network algorithm was applied to each data set on Landsat imagery scene. It was possible to generate an mean air temperature map from June to September 2019 and confirmed that detailed air temperature information could be estimated. The result is expected to be utilized for national disaster safety management such as heat wave response policies and heat island mitigation research.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.21 no.5
• /
• pp.599-607
• /
• 2015

The purpose of this study is to reproduce long-term wave fields in coastal waters of Korea based on wave hind-casting modelling and discuss its applications. To validate wind data(NCEP, ECMWF, JMA-MSM), comparison of wind data was done with wave buoy data. JMA-MSM predicted wind data with high accuracy. But due to relatively longer period of ECMWF wind data as compared to that of JMA-MSM, wind data set of ECMWF(2001~2014) was used to perform wave hind-casting modelling. Results from numerical modelling were verified with the observed data of wave buoys installed by Korea Meteorological Administration(KMA) and Korea Hydrographic and Oceanographic Agency(KHOA) on offshore waters. The results agree well with observations at buoy stations, especially during the event periods such as a typhoon. Consequently, the wave data reproduced by wave hind-casting modelling was used to obtain missing data in wave observation buoys. The obtained missing data indicated underestimation of maximum wave height during the event period at some points of buoys. Reasons for such underestimation may be due to larger time interval and resolution of the input wind data, water depth and grid size etc. The methodology used in present study can be used to analyze coastal erosion data in conjunction with a wave characteristic of the event period in coastal areas. Additionally, the method can be used in the coastal disaster vulnerability assessment to generate wave points of interest.

• Journal of the Korean Society of Food Culture
• /
• v.16 no.5
• /
• pp.407-415
• /
• 2001

The purposes of this study were to assess th importance and performance of food suppliers and to explore the ways to improve the school foodservice purchasing management. The questionnaire regarding purchasing was composed of two parts. The part one consisted of questions on general characteristics of dieticians and school foodservice operations, and the part two was composed of questions on the importance & performance of food suppliers. Completed questionnaires were received from 286 dieticians of elementary school foodservice operations in Seoul. Statistical data analysis was completed using the SPSS/win for descriptive and t-test. In dieticians' demographic data, 36.6% were over 30 years old, 32.4% were $28{\sim}29$ years of age and 31.0% were below 27years old. Most of the respondents(68.3%) had overall working experiences less than 58 months and almost half of them(56.3%) were married. The school foodservice operations which employed a chef were 50.3%. The food suppliers' attributes that were recognized by the dieticians with high value of mean importance scores were: food quality, maintenance of food quality, accuracy in filling orders, quality of delivery facilities, on time delivery and packaging. Average mean score for importance was 4.33('important') out of 5 and mean score for performance was 3.50('so-so') out of 5. By IPA techniques, the attributes that deserve higher attention were those that ranked high in importance and low in performance(Quadrant A). The coordinates in Quadrant A were geographical location of supplier, maintenance of food quality, provision of information on cost variation and salesman's knowledge.

• Bulletin of the Korean Chemical Society
• /
• v.35 no.7
• /
• pp.1944-1950
• /
• 2014

Employment of a time consuming, sophisticated calculation using the all-atom force field and generalized-Born implicit solvent model (GBIS) for refinement of NMR structures has become practical through advances in computational methods and capacities. GBIS refinement improves the qualities of the resulting NMR structures with reduced computational times. However, the contribution of GBIS to NMR structures has not been sufficiently studied in a quantitative way. In this paper, we report the effects of GBIS on the refined NMR structures of ubiquitin (UBQ) and GB1 with subsets of distance restraints derived from experimental data. Random omission prepared a series of distance restraints 0.05, 0.1, 0.3, 0.5, and 0.7 times smaller. For each number, we produced five different restraints for statistical analysis. We then recalculated the NMR structures using CYANA software, followed by GBIS refinements using the AMBER package. GBIS improved both the precision and accuracy of all the structures, but to varied levels. The degrees of improvement were significant when the input restraints were insufficient. In particular, GBIS enabled GB1 to form an accurate structure even with distance restraints of 5%, revealing that the root-mean-square deviation was less than 1 ${\AA}$ from the X-ray backbone structure. We also showed that the efficiency of searching the conformational space was more important for finding accurate structures with the calculation of UBQ with 5% distance restraints than the number of conformations generated. Our data will provide a meaningful guideline to judge and compare the structural improvements by GBIS.

• Korean Journal of Remote Sensing
• /
• v.26 no.5
• /
• pp.537-547
• /
• 2010

The importance of vegetation in studies of global climate and biogeochemical cycles is well recognized. Especially. the FPAR (fraction of photosynthetically active radiation) is one of the important parameters in ecosystem productivity and carbon budget models. Therefore, accurate estimates of vegetation parameters are increasingly important in environmental impact assessment studies. In this study, optical FPAR using the Terra MODIS (MODerate resolution Imaging Spectroradiometer), SPOT VEGETATION and ECOCLIMAP data reproduced on the Korean peninsula. We applied the empirical method which is usually estimated as a linear or nonlinear function of vegetation indices. As results, we estimated the accurate expression which is 0.9039 of $R^2$ in cropland and 0.7901 of $R^2$ in forest. Finally, this study could be demonstrated to calibrate that produced FPAR while the overall pattern and random noise through the comparative analysis of FPAR on the reference data. Optimal use of input parameter on the Korean peninsula should be helping the accuracy of output as well as the improved quality of research.

• Dementia and Neurocognitive Disorders
• /
• v.23 no.3
• /
• pp.146-160
• /
• 2024

Background and Purpose: The emotions of people at various stages of dementia need to be effectively utilized for prevention, early intervention, and care planning. With technology available for understanding and addressing the emotional needs of people, this study aims to develop speech emotion recognition (SER) technology to classify emotions for people at high risk of dementia. Methods: Speech samples from people at high risk of dementia were categorized into distinct emotions via human auditory assessment, the outcomes of which were annotated for guided deep-learning method. The architecture incorporated convolutional neural network, long short-term memory, attention layers, and Wav2Vec2, a novel feature extractor to develop automated speech-emotion recognition. Results: Twenty-seven kinds of Emotions were found in the speech of the participants. These emotions were grouped into 6 detailed emotions: happiness, interest, sadness, frustration, anger, and neutrality, and further into 3 basic emotions: positive, negative, and neutral. To improve algorithmic performance, multiple learning approaches were applied using different data sources-voice and text-and varying the number of emotions. Ultimately, a 2-stage algorithm-initial text-based classification followed by voice-based analysis-achieved the highest accuracy, reaching 70%. Conclusions: The diverse emotions identified in this study were attributed to the characteristics of the participants and the method of data collection. The speech of people at high risk of dementia to companion robots also explains the relatively low performance of the SER algorithm. Accordingly, this study suggests the systematic and comprehensive construction of a dataset from people with dementia.

• The Korean Journal of Food And Nutrition
• /
• v.28 no.5
• /
• pp.871-879
• /
• 2015

24-hour recall is the dietary assessment method most frequently used to evaluate dietary intake; however, accuracy is an issue when using this method, especially in large-scale studies. The purpose of this study was to assess the validity of dietary intake estimation using one serving size. Estimates of energy and nutrients taken in over a 24-hr period based on actual intake amount (24HRAI) and based on estimates of one serving size (24HRSS) were compared. Data were analyzed using a paired t-test, Pearson's correlation coefficients, and a cross-classification method. In male subjects, intake levels of energy, fat, vitamin C, vitamin $B_1$, Zn, and total food measured using 24HRAI were significantly higher than those measured using 24HRSS. In female subjects, intake of carbohydrates, fiber, fat, vitamin A, vitamin C, vitamin B complexes, various minerals, and total food measured using 24HRAI were significantly lower than those measured using 24HRSS. Energy-adjusted Pearson's correlation coefficients revealed that intake of all nutrients showed a significant positive relationship between the two measurement methods in both males and females. Cross-classification analysis revealed that 50.5~67.6% of women and 40.3~71% of men were classified in the same quartile of intake of each nutrient when comparing data from 24HRAI and 24HRSS. We conclude that using one serving size in 24-hr recall analysis was valid and therefore may be used in studies to assess food consumption in the general adult population. Also, this method can be used to classify energy and nutrient intake into quartile, which is useful in examining the association between diet and chronic diseases.