• The KIPS Transactions:PartB
• /
• v.12B no.4 s.100
• /
• pp.487-498
• /
• 2005

Recently, It has focused on decision tree algorithm that can handle large dataset. However, because most of these algorithms for large datasets process data in a batch mode, if new data is added, they have to rebuild the tree from scratch. h more efficient approach to reducing the cost problem of rebuilding is an approach that builds a tree incrementally. Representative algorithms for incremental tree construction methods are BOAT and ITI and most of these algorithms use a local discretization method to handle the numeric data type. However, because a discretization requires sorted numeric data in situation of processing large data sets, a global discretization method that sorts all data only once is more suitable than a local discretization method that sorts in every node. This paper proposes an incremental tree construction method that efficiently rebuilds a tree using a global discretization method to handle the numeric data type. When new data is added, new categories influenced by the data should be recreated, and then the tree structure should be changed in accordance with category changes. This paper proposes a method that extracts sample points and performs discretiration from these sample points to recreate categories efficiently and uses confidence intervals and a tree restructuring method to adjust tree structure to category changes. In this study, an experiment using people database was made to compare the proposed method with the existing one that uses a local discretization.

• Journal of the Korean Society for Railway
• /
• v.19 no.3
• /
• pp.314-323
• /
• 2016

Overhead line voltage of DC railway traction substations has rising or falling characteristics depending on the acceleration and regenerative braking of the subway train loads. The suppression of this irregular fluctuation of the line voltage gives rise to improved energy efficiency of both the railway substation and the trains. This paper presents parameter estimation schemes using the RC circuit model for an overhead line voltage at a 1500V DC electric railway traction substation. A linear artificial neural network with a back-propagation learning algorithm was trained using the measurement data for an overhead line voltage and four feeder currents. The least square estimation method was configured to implement batch processing of these measurement data. These estimation results have been presented and performance analysis has been achieved through raw data simulation.

• Journal of Power Electronics
• /
• v.13 no.4
• /
• pp.516-527
• /
• 2013

This paper presents a new overall system for state-of-available-power (SoAP) prediction for a lithium-ion battery pack. The essential part of this method is based on an adaptive network architecture which utilizes both fuzzy model (FIS) and artificial neural network (ANN) into the framework of adaptive neuro-fuzzy inference system (ANFIS). While battery aging proceeds, the system is capable of delivering accurate power prediction not only for room temperature, but also at lower temperatures at which power prediction is most challenging. Due to design property of ANN, the network parameters are adapted on-line to the current battery states (state-of-charge (SoC), state-of-health (SoH), temperature). SoC is required as an input parameter to SoAP module and high accuracy is crucial for a reliable on-line adaptation. Therefore, a reasonable way to determine the battery state variables is proposed applying a combination of several partly different algorithms. Among other SoC boundary estimation methods, robust extended Kalman filter (REKF) for recalibration of amp hour counters was implemented. ANFIS then achieves the SoAP estimation by means of time forward voltage prognosis (TFVP) before a power pulse occurs. The trade-off between computational cost of batch-learning and accuracy during on-line adaptation was optimized resulting in a real-time system with TFVP absolute error less than 1%. The verification was performed on a software-in-the-loop test bench setup using a 53 Ah lithium-ion cell.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.9
• /
• pp.1731-1737
• /
• 2017

K Nearest Neighbor Join (KNN Join) is a simple yet effective method in machine learning. It is widely used in small dataset of the past time. As the number of data increases, it is infeasible to run this model on an actual application by a single machine due to memory and time restrictions. Nowadays a popular batch process model called MapReduce which can run on a cluster with a large number of computers is widely used for large-scale data processing. Hadoop is a framework to implement MapReduce, but its performance can be further improved by a new framework named Spark. In the present study, we will provide a KNN Join implement based on Spark. With the advantage of its in-memory calculation capability, it will be faster and more effective than Hadoop. In our experiments, we study the influence of different factors on running time and demonstrate robustness and efficiency of our approach.

• Journal of Broadcast Engineering
• /
• v.27 no.6
• /
• pp.872-884
• /
• 2022

With the development of the IoT industry, different types of time series data are being generated in various industries, and it is evolving into research that reproduces and utilizes it through re-integration. In addition, due to data processing speed and issues of the utilization system in the actual industry, there is a growing tendency to compress the size of data when using time series data and integrate it. However, since the guidelines for integrating time series data are not clear and each characteristic such as data description time interval and time section is different, it is difficult to use it after batch integration. In this paper, two integration methods are proposed based on the integration criteria setting method and the problems that arise during integration of time series data. Based on this, integration framework of a heterogeneous time series data was constructed that is considered the characteristics of time series data, and it was confirmed that different heterogeneous time series data compressed can be used for integration and various machine learning.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• Geomechanics and Engineering
• /
• v.35 no.2
• /
• pp.121-133
• /
• 2023

The demand for cement and limestone crushed materials has increased many folds due to the tremendous increase in construction activities in Pakistan during the past few decades. The number of cement production industries has increased correspondingly, and so the rock-blasting operations at the limestone quarry sites. However, the safety procedures warranted at these sites for the blast-induced ground vibrations (BIGV) have not been adequately developed and/or implemented. Proper prediction and monitoring of BIGV are necessary to ensure the safety of structures in the vicinity of these quarry sites. In this paper, an attempt has been made to predict BIGV using artificial neural network (ANN) at three selected limestone quarries of Pakistan. The ANN has been developed in Python using Keras with sequential model and dense layers. The hyper parameters and neurons in each of the activation layers has been optimized using randomized and grid search method. The input parameters for the model include distance, a maximum charge per delay (MCPD), depth of hole, burden, spacing, and number of blast holes, whereas, peak particle velocity (PPV) is taken as the only output parameter. A total of 110 blast vibrations datasets were recorded from three different limestone quarries. The dataset has been divided into 85% for neural network training, and 15% for testing of the network. A five-layer ANN is trained with Rectified Linear Unit (ReLU) activation function, Adam optimization algorithm with a learning rate of 0.001, and batch size of 32 with the topology of 6-32-32-256-1. The blast datasets were utilized to compare the performance of ANN, multivariate regression analysis (MVRA), and empirical predictors. The performance was evaluated using the coefficient of determination (R²), mean absolute error (MAE), mean squared error (MSE), mean absolute percentage error (MAPE), and root mean squared error (RMSE)for predicted and measured PPV. To determine the relative influence of each parameter on the PPV, sensitivity analyses were performed for all input parameters. The analyses reveal that ANN performs superior than MVRA and other empirical predictors, andthat83% PPV is affected by distance and MCPD while hole depth, number of blast holes, burden and spacing contribute for the remaining 17%. This research provides valuable insights into improving safety measures and ensuring the structural integrity of buildings near limestone quarry sites.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.199-207
• /
• 2024

Local reservoirs are crucial sources for agricultural water supply, necessitating stable water level management to prepare for extreme climate conditions such as droughts. Water level prediction is significantly influenced by local climate characteristics, such as localized rainfall, as well as seasonal factors including cropping times, making it essential to understand the correlation between input and output data as much as selecting an appropriate prediction model. In this study, extensive multivariate data from over 400 reservoirs in Jeollabuk-do from 1991 to 2022 was utilized to train and validate a water level prediction model that comprehensively reflects the complex hydrological and climatological environmental factors of each reservoir, and to analyze the impact of each input feature on the prediction performance of water levels. Instead of focusing on improvements in water level performance through neural network structures, the study adopts a basic Feedforward Neural Network composed of fully connected layers, batch normalization, dropout, and activation functions, focusing on the correlation between multivariate input data and prediction performance. Additionally, most existing studies only present short-term prediction performance on a daily basis, which is not suitable for practical environments that require medium to long-term predictions, such as 10 days or a month. Therefore, this study measured the water level prediction performance up to one month ahead through a recursive method that uses daily prediction values as the next input. The experiment identified performance changes according to the prediction period and analyzed the impact of each input feature on the overall performance based on an Ablation study.