• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.6
• /
• pp.1530-1544
• /
• 2023

Data monitoring is an important foundation of modern science. In most cases, the monitoring data is time-series data, which has high application value. The deep learning algorithm has a strong nonlinear fitting capability, which enables the recognition of time series by capturing anomalous information in time series. At present, the research of time series recognition based on deep learning is especially important for data monitoring. Deep learning algorithms require a large amount of data for training. However, abnormal sample is a small sample in time series, which means the number of abnormal time series can seriously affect the accuracy of recognition algorithm because of class imbalance. In order to increase the number of abnormal sample, a data augmentation method called GANBATS (GAN-based Bi-LSTM and Attention for Time Series) is proposed. In GANBATS, Bi-LSTM is introduced to extract the timing features and then transfer features to the generator network of GANBATS.GANBATS also modifies the discriminator network by adding an attention mechanism to achieve global attention for time series. At the end of discriminator, GANBATS is adding averagepooling layer, which merges temporal features to boost the operational efficiency. In this paper, four time series datasets and five data augmentation algorithms are used for comparison experiments. The generated data are measured by PRD(Percent Root Mean Square Difference) and DTW(Dynamic Time Warping). The experimental results show that GANBATS reduces up to 26.22 in PRD metric and 9.45 in DTW metric. In addition, this paper uses different algorithms to reconstruct the datasets and compare them by classification accuracy. The classification accuracy is improved by 6.44%-12.96% on four time series datasets.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.338-341
• /
• 2003

Recently, the chaotic method is employed to forecast a near future of uncertain phenomena. This method makes it possible by restructuring an attractor of given time-series data in multi-dimensional space through Takens' embedding theory. However, many economical time-series data are not sufficiently chaotic. In other words, it is hard to forecast the future trend of such economical data on the basis of chaotic theory. In this paper, time-series data are divided into wave components using wavelet transform. It is shown that some divided components of time-series data show much more chaotic in the sense of correlation dimension than the original time-series data. The highly chaotic nature of the divided component enables us to precisely forecast the value or the movement of the time-series data in near future. The up and down movement of TOPICS value is shown so highly predicted by this method as 70％.

• Journal of the Korean Data and Information Science Society
• /
• v.14 no.2
• /
• pp.265-274
• /
• 2003

Time series data are influenced by the external events such as holiday, strike, oil shock, and political change, so the external events cause a sudden change to the time series data. We regard the observation as outlier that occurred as a result of external events. In general, it is called intervention if we know the period and the reason of external events, and it makes an analyst difficult to establish a time series model. Therefore, it is important that we analyze the styles and effects of intervention. In this paper, we considered the linear time series model with invention and compared with nonlinear time series models such as ARCH, GARCH model and also we compared with the combination prediction method that Tong(1990) introduced. In the practical case study, we compared prediction power with RMSE among linear, nonlinear time series model with intervention and combination prediction method.

• 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.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1996.10a
• /
• pp.294-297
• /
• 1996

This paper proposes the method to recognize of time series data based on the chaotic feature extraction. Features extract from time series data using the chaotic time series data analysis and the pattern recognition process is using a neural network classifier. In experiment, EEG(electroencephalograph) signals are extracted features by correlation dimension and Lyapunov experiments, and these features are classified by multilayer perceptron neural networks. Proposed chaotic feature extraction enhances recognition results from chaotic time series data.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.50 no.5
• /
• pp.233-240
• /
• 2001

This paper presents a time series prediction method using a fuzzy rule-based system. Extracting fuzzy rules by performing a simple one-pass operation on the training data is quite attractive because it is easy to understand, verify, and extend. The simplest method is probably to relate an estimate, x(n+k), with past data such as x(n), x(n-1), ..x(n-m), where k and m are prefixed positive integers. The relation is represented by fuzzy if-then rules, where the past data stand for premise part and the predicted value for consequence part. However, a serious problem of the method is that it cannot handle nonstationary data whose long-term mean is varying. To cope with this, a new training method is proposed, which utilizes the difference of consecutive data in a time series. In this paper, typical previous works relating time series prediction are briefly surveyed and a new method is proposed to overcome the difficulty of prediction nonstationary data. Finally, computer simulations are illustrated to show the improved results for various time series.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.210-212
• /
• 2022

In the manufacturing industry, various sensors are attached to monitor the status of production facility. In many cases, the data obtained through these sensors is time series data. In order to determine whether the status of the production facility is abnormal, the process of extracting patterns from time series data must be preceded. Also various methods for extracting patterns from time series data are studied. In this paper, we use matrix profile algorithm to extract patterns from the collected multivariate time series data. Through this, the pattern of multi sensor data currently being collected from the CNC machine is extracted.

• Communications for Statistical Applications and Methods
• /
• v.22 no.3
• /
• pp.209-222
• /
• 2015

Time series are used in many studies for model building and analysis. We must be very careful to understand the kind of time series data used in the analysis. In this review article, we will begin with some issues related to the use of aggregate and systematic sampling time series. Since several time series are often used in a study of the relationship of variables, we will also consider vector time series modeling and analysis. Although the basic procedures of model building between univariate time series and vector time series are the same, there are some important phenomena which are unique to vector time series. Therefore, we will also discuss some issues related to vector time models. Understanding these issues is important when we use time series data in modeling and analysis, regardless of whether it is a univariate or multivariate time series.

• Communications for Statistical Applications and Methods
• /
• v.21 no.6
• /
• pp.471-486
• /
• 2014

This study introduces a new type of symbolic data, a candle chart-valued time series. We aggregate four stock indices (i.e., open, close, highest and lowest) as a one data point to summarize a huge amount of data. In other words, we consider a candle chart, which is constructed by open, close, highest and lowest stock indices, as a type of symbolic data for a long period. The proposed candle chart-valued time series effectively summarize and visualize a huge data set of stock indices to easily understand a change in stock indices. We also propose novel approaches for the candle chart-valued time series modeling based on a combination of two midpoints and two half ranges between the highest and the lowest indices, and between the open and the close indices. Furthermore, we propose three types of sum of square for estimation of the candle chart valued-time series model. The proposed methods take into account of information from not only ordinary data, but also from interval of object, and thus can effectively perform for time series modeling (e.g., forecasting future stock index). To evaluate the proposed methods, we describe real data analysis consisting of the stock market indices of five major Asian countries'. We can see thorough the results that the proposed approaches outperform for forecasting future stock indices compared with classical data analysis.

• Journal of Korean Institute of Industrial Engineers
• /
• v.25 no.1
• /
• pp.111-124
• /
• 1999

This evapaper is toluate the forecasting performance of three neural network(NN) approaches against ARIMA model using the famous time series analysis competition data. The first NN approach is to analyze the second Makridakis (M2) Competition Data using Multilayer Perceptron (MLP) that has been the most popular NN model in time series analysis. Since it is recently known that MLP suffers from bias/variance dilemma, two approaches are suggested in this study. The second approach adopts Cascade Correlation Network (CCN) that was suggested by Fahlman & Lebiere as an alternative to MLP. In the third approach, a time series is separated into two series using Noise Filtering Network (NFN) that utilizes autoassociative memory function of neural network. The forecasts in the decomposition analysis are the sum of two prediction values obtained from modeling each decomposed series, respectively. Among the three NN approaches, Decomposition Analysis shows the best forecasting performance on the M2 Competition Data, and is expected to be a promising tool in analyzing socio-economic time series data because it reduces the effect of noise or outliers that is an impediment to modeling the time series generating process.