• Journal of Korean Society for Quality Management
• /
• v.48 no.3
• /
• pp.511-520
• /
• 2020

Purpose: The purpose of this study was to devise an accurate machine learning model for predicting silica concentrations following the addition of impurities, through time series analysis of mining data. Methods: The mining data were preprocessed and subjected to time series analysis using the machine learning model. Through correlation analysis, valid variables were selected and meaningless variables were excluded. To reflect changes over time, dependent variables at baseline were treated as independent variables at later time points. The relationship between independent variables and the dependent variable after n point was subjected to Pearson correlation analysis. Results: The correlation (R2) was strongest after 3 hours, which was adopted as a dependent variable. According to root mean square error (RMSE) data, the proposed method was superior to the other machine learning methods. The XGboost algorithm showed the best predictive performance. Conclusion: This study is important given the current lack of machine learning studies pertaining to the domestic mining industry. In addition, using time series analysis in mining data will show further improvement. Before establishing a predictive model for the proposed method, predictions should be made using data with time series characteristics. After doing this work, it should also improve prediction accuracy in other domains.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.5 no.5
• /
• pp.419-426
• /
• 2000

Generally diode rectifier has been used as Preregulator of arc- welding machine. Its defect is the generation of high current harmonics which result In utility pollution. In this parer, using the three-phase PWM converter, the input performance of arc-welding machine is improved by increasing the input power factor. When the diode rectifier in the arc-welding machine is replaced with three-phase PWM converter, PWM converter should control DC voltage of the output side. If PI controller is used for the output DC voltage regulator, the output DC voltage has high ripple voltage due to the instantaneous load variation in the arc-welding machine. In this paper, the feedforward compensator has been introduced to reduce the DC voltage ripple. It lessens the influence of load current which is the disturbance of the output DC voltage regulator. Theoretical prediction of this analysis has been verified by comparing with experimental data

• Journal of Aerospace System Engineering
• /
• v.13 no.5
• /
• pp.65-71
• /
• 2019

In this study, as a part of the aging aircraft performance improvement project for which no design information is provided, a new type of blade antenna is installed on the main part of the aging aircraft, and the method of proving the fatigue life of the main part of the aircraft is reviewed and summarized. There are various methods to prove fatigue life according to the manufacturer and aircraft design conditions. The fatigue life prediction and damage tolerance range of the relevant site were obtained through related regulations and industry examples. From these results, the fatigue life of newly installed antennas around the main parts of the aging aircraft was evaluated and the maintenance period and criteria were set according to the damage tolerance.

• The Journal of the Korea Contents Association
• /
• v.7 no.8
• /
• pp.13-20
• /
• 2007

Decision tree is one of the most useful analysis methods for various data mining functions, including prediction, classification, etc, from massive data. Decision tree grows by splitting nodes, during which the purity increases. It is needed to stop splitting nodes when the purity does not increase effectively or new leaves does not contain meaningful number of records. Pruning is done if a branch does not show certain level of performance. By pruning, the structure of decision tree is changed and it is implied that the previous splitting of the parent node was not effective. It is also implied that the splitting of the ancestor nodes were not effective and the choices of attributes and criteria in splitting them were not successful. It should be noticed that new attributes or criteria might be selected to split such nodes for better tries. In this paper, we suggest a procedure to modify decision tree by Fuzzy theory and splitting as an integrated approach.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.7
• /
• pp.1707-1717
• /
• 1998

Polyphase filtering techniques is used to suppress the narrowband jammer signal such as USDC TDMA overlaying the band occupied by DS-CDMA system. In the proposed jammer suppression, the received signal is separated into 64 subchannels in two stages by polyphase filtering and the location of the narrowband jammer signal is determined by measuring each subchannel power and the contaminated subchannels are simply blocked. The $E_{b}/N_{0}$ 0/ improvement of the CDMA system from jammer suppession was outstanding. The $E_{b}/N_{0}$ degradation in comparison with a performance of no jammer is around 0.8dB in the worst case. The results are also compared with those of linear prediction jammer suppression. The implementation of the ployphase jammer suppression requires great amount of data processing and computation compared to linear predication filter. Thus it is more appropriate to implement with a ASIC rather than WITH several DSPs for user terminals of forward link.

• Journal of the Korean Geotechnical Society
• /
• v.24 no.9
• /
• pp.33-40
• /
• 2008

This paper describes the performance of ground improvement project using prefabricated vertical drains of condition, in which approximately 10m dredged fill overlies original soft foundation layer in the coastal area composed of soft marine clay with high water content and high compressibility. From field monitoring results, excessive ground settlement compared with predicted settlement in design stage developed during the following one year. In order to predict the final consolidation behavior, recalculation of consolidation settlements and back analysis using observed settlements were conducted. Field monitoring results of surface settlements were evaluated, and then corrected because large shear deformation occurred by construction events in the early stages of consolidation. To predict the consolidation behavior, material functions and in-situ conditions from laboratory consolidation test were re-analyzed. Using these results, height of additional embankment is estimated to satisfy residual settlement limit and maintain an adequate ground elevation. The recalculated time-settlement curve has been compared with field monitoring results after additional surcharge was applied. It might be used for verification of recalculated results.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.334-334
• /
• 2022

Improvement of old-fashioned rain gauge systems for automatic, timely, continuous, and accurate precipitation observation is highly essential for weather/climate prediction and natural hazards early warning, since the occurrence frequency and intensity of heavy and extreme precipitation events (especially floods) are recently getting more increase and severe worldwide due to climate change. Although rain gauge accuracy of 0.1 mm is recommended by the World Meteorological Organization (WMO), the traditional rain gauges in both weighting and tipping bucket types are often unable to meet that demand due to several existing technical limitations together with higher production and maintenance costs. Therefore, we aim to introduce a newly developed and cost-effective hybrid rain gauge system at 0.1 mm accuracy that combines advantages of weighting and tipping bucket types for continuous, automatic, and accurate precipitation observation, where the errors from long-term load cells and external environmental sources (e.g., winds) can be removed via an automatic drainage system and artificial intelligence-based data quality control procedure. Our rain gauge system consists of an instrument unit for measuring precipitation, a communication unit for transmitting and receiving measured precipitation signals, and a database unit for storing, processing, and analyzing precipitation data. This newly developed rain gauge was designed according to the weather instrument criteria, where precipitation amounts filled into the tipping bucket are measured considering the receiver's diameter, the maximum measurement of precipitation, drainage time, and the conductivity marking. Moreover, it is also designed to transmit the measured precipitation data stored in the PCB through RS232, RS485, and TCP/IP, together with connecting to the data logger to enable data collection and analysis based on user needs. Preliminary results from a comparison with an existing 1.0-mm tipping bucket rain gauge indicated that our developed rain gauge has an excellent performance in continuous precipitation observation with higher measurement accuracy, more correct precipitation days observed (120 days), and a lower error of roughly 27 mm occurred during the measurement period.

• Journal of Broadcast Engineering
• /
• v.25 no.3
• /
• pp.338-345
• /
• 2020

Versatile Video Coding (VVC), which has been developing as a next generation video coding standard, has adopted various techniques to achieve more than twice the compression performance of HEVC (High Efficiency Video Coding). The recently released VVC Test Model (VTM) shows 38% Bjontegaard Delta bitrate (BD-rate) improvement and 9x/1.6x encoding/decoding complexity over HEVC. In order to reduce such increased complexity, various fast algorithms have been proposed. In this paper, gradient-based methods of fast intra mode decision and block splitting are presented. Experimental results show that, compared to VTM6.0, the proposed method gives up to 65% encoding time reduction with 3.54% BD-rate loss in All-Intra (AI) configuration.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.24 no.6
• /
• pp.684-693
• /
• 2016

The luxury car industry has grown 10.5 % every year from 2010 to 2014. For this reason, it is very important for automotive companies to improve profitability and brand value. High-performance brake systems have become an absolute necessity because of the increase in engine power and customer preference among other factors. Also, competing automotive companies actively reinforce domestic production in order to maintain quality and infrastructure for luxury cars. In this regard, we demonstrated new carbon ceramic brakes to improve brake reliability for luxury cars and to improve the competitiveness of automotive companies. Finally, we propose the next-generation braking technology by predicting technological evolution trends.

• Journal of KIISE:Databases
• /
• v.36 no.4
• /
• pp.247-253
• /
• 2009

Rank learning and processing have gained much attention in the IR and data mining communities for the last decade. While other data mining techniques such as classification and regression have been actively researched to interoperate with RDBMS by using the tightly coupled or loose coupling approaches, ranking has been researched independently without integrating into RDBMS. This paper proposes a tightly coupled integration of the Ranking SVM into MySQL in order to perform the rank learning task efficiently within the RDBMS. We implemented new SQL commands for learning ranking functions and predicting ranking scores. We evaluated our tightly coupled integration of Ranking SVM by comparing it to a loose coupling implementation. The experiment results show that our approach has a performance improvement of $10{\sim}40%$ in the training phase and 60% in the prediction phase.