Medical Time-Series Data Generation Using Generative Adversarial Networks

As a measure of the brain's electrical activity, electroencephalography (EEG) is the primary signal of interest for brain-computer-interfaces (BCI). BCIs offer a communication pathway between a brain and an external device, translating thought into action with suitable processing. EEG data is the most common signal source for such technologies. However, artefacts induced in BCIs in the real-world context can severely degrade their performance relative to their in-laboratory performance. In most cases, the recorded signals are so heavily corrupted by noise that they are unusable and restrict BCI's broader applicability. To realise the use of portable BCIs capable of high-quality performance in a real-world setting, we use Generative Adversarial Networks (GANs) that can adopt both supervised and unsupervised learning approaches. Although our approach is supervised, the same model can be used for unsupervised tasks such as data augmentation/imputation in the low resource setting. Exploiting recent advancements in Generative Adversarial Networks (GAN), we construct a pipeline capable of denoising artefacts from EEG time series data. In the case of denoising data, it maps noisy EEG signals to clean EEG signals, given the nature of the respective artefact. We demonstrate the capability of our network on a toy dataset and a benchmark EEG dataset developed explicitly for deep learning denoising techniques. Our datasets consist of an artificially added mains noise (50/60 Hz) artefact dataset and an open-source EEG benchmark dataset with two artificially added artefacts. Artificially inducing myogenic and ocular artefacts for the benchmark dataset allows us to present qualitative and quantitative evidence of the GANs denoising capabilities and rank it among the current gold standard deep learning EEG denoising techniques. We show the power spectral density (PSD), signal-to-noise ratio (SNR), and other classical time series similarity measures for quantitative metrics and compare our model to those previously used in the literature. To our knowledge, this framework is the first example of a GAN capable of EEG artefact removal and generalisable to more than one artefact type. Our model has provided a competitive performance in advancing the state-of-the-art deep learning EEG denoising techniques. Furthermore, given the integration of AI into wearable technology, our method would allow for portable EEG devices with less noisy and more stable brain signals.

Download Full-text

MAD-GAN: Multivariate Anomaly Detection for Time Series Data with Generative Adversarial Networks

Artificial Neural Networks and Machine Learning – ICANN 2019: Text and Time Series - Lecture Notes in Computer Science ◽

10.1007/978-3-030-30490-4_56 ◽

2019 ◽

pp. 703-716 ◽

Cited By ~ 26

Author(s):

Dan Li ◽

Dacheng Chen ◽

Baihong Jin ◽

Lei Shi ◽

Jonathan Goh ◽

...

Keyword(s):

Time Series ◽

Anomaly Detection ◽

Time Series Data ◽

Series Data ◽

Generative Adversarial Networks ◽

Adversarial Networks

Download Full-text

..Dis-AE-LSTM: Generative Adversarial Networks for Anomaly Detection of Time Series Data

2020 International Conference on Artificial Intelligence and Computer Engineering (ICAICE) ◽

10.1109/icaice51518.2020.00070 ◽

2020 ◽

Author(s):

Sheng Mao ◽

Jiansheng Guo ◽

Taoyong Gu ◽

Zhong Ma

Keyword(s):

Time Series ◽

Anomaly Detection ◽

Time Series Data ◽

Series Data ◽

Generative Adversarial Networks ◽

Adversarial Networks

Download Full-text

A Time-Series Data Generation Method to Predict Remaining Useful Life

Processes ◽

10.3390/pr9071115 ◽

2021 ◽

Vol 9 (7) ◽

pp. 1115

Author(s):

Gilseung Ahn ◽

Hyungseok Yun ◽

Sun Hur ◽

Si-Yeong Lim

Keyword(s):

Time Series ◽

Time Series Data ◽

Remaining Useful Life ◽

Series Data ◽

Generation Model ◽

Data Generation ◽

Training Time ◽

Symbolic Aggregate Approximation ◽

Useful Life ◽

Occurrence Patterns

Accurate predictions of remaining useful life (RUL) of equipment using machine learning (ML) or deep learning (DL) models that collect data until the equipment fails are crucial for maintenance scheduling. Because the data are unavailable until the equipment fails, collecting sufficient data to train a model without overfitting can be challenging. Here, we propose a method of generating time-series data for RUL models to resolve the problems posed by insufficient data. The proposed method converts every training time series into a sequence of alphabetical strings by symbolic aggregate approximation and identifies occurrence patterns in the converted sequences. The method then generates a new sequence and inversely transforms it to a new time series. Experiments with various RUL prediction datasets and ML/DL models verified that the proposed data-generation model can help avoid overfitting in RUL prediction model.

Download Full-text

A forecasting of stock trading price using time series information based on big data

International Journal of Electrical and Computer Engineering (IJECE) ◽

10.11591/ijece.v11i3.pp2548-2554 ◽

2021 ◽

Vol 11 (3) ◽

pp. 2548

Author(s):

Soo-Tai Nam ◽

Chan-Yong Jin ◽

Seong-Yoon Shin

Keyword(s):

Time Series ◽

Big Data ◽

Stock Price ◽

Euclidean Distance ◽

Time Series Data ◽

Series Data ◽

Analysis Tool ◽

Large Set ◽

Data Generation ◽

Management Tools

Big data is a large set of structured or unstructured data that can collect, store, manage, and analyze data with existing database management tools. And it means the technique of extracting value from these data and interpreting the results. Big data has three characteristics: The size of existing data and other data (volume), the speed of data generation (velocity), and the variety of information forms (variety). The time series data are obtained by collecting and recording the data generated in accordance with the flow of time. If the analysis of these time series data, found the characteristics of the data implies that feature helps to understand and analyze time series data. The concept of distance is the simplest and the most obvious in dealing with the similarities between objects. The commonly used and widely known method for measuring distance is the Euclidean distance. This study is the result of analyzing the similarity of stock price flow using 793,800 closing prices of 1,323 companies in Korea. Visual studio and Excel presented calculate the Euclidean distance using an analysis tool. We selected “000100” as a target domestic company and prepared for big data analysis. As a result of the analysis, the shortest Euclidean distance is the code “143860” company, and the calculated value is “11.147”. Therefore, based on the results of the analysis, the limitations of the study and theoretical implications are suggested.

Download Full-text

Generating Energy Data for Machine Learning with Recurrent Generative Adversarial Networks

Energies ◽

10.3390/en13010130 ◽

2019 ◽

Vol 13 (1) ◽

pp. 130 ◽

Cited By ~ 4

Author(s):

Mohammad Navid Fekri ◽

Ananda Mohon Ghosh ◽

Katarina Grolinger

Keyword(s):

Machine Learning ◽

Neural Networks ◽

Energy Consumption ◽

Smart Grid ◽

Time Series Data ◽

Security And Privacy ◽

Series Data ◽

Generative Adversarial Networks ◽

Generative Adversarial Network ◽

Adversarial Networks

The smart grid employs computing and communication technologies to embed intelligence into the power grid and, consequently, make the grid more efficient. Machine learning (ML) has been applied for tasks that are important for smart grid operation including energy consumption and generation forecasting, anomaly detection, and state estimation. These ML solutions commonly require sufficient historical data; however, this data is often not readily available because of reasons such as data collection costs and concerns regarding security and privacy. This paper introduces a recurrent generative adversarial network (R-GAN) for generating realistic energy consumption data by learning from real data. Generativea adversarial networks (GANs) have been mostly used for image tasks (e.g., image generation, super-resolution), but here they are used with time series data. Convolutional neural networks (CNNs) from image GANs are replaced with recurrent neural networks (RNNs) because of RNN’s ability to capture temporal dependencies. To improve training stability and increase quality of generated data, Wasserstein GANs (WGANs) and Metropolis-Hastings GAN (MH-GAN) approaches were applied. The accuracy is further improved by adding features created with ARIMA and Fourier transform. Experiments demonstrate that data generated by R-GAN can be used for training energy forecasting models.

Download Full-text

A Comparative Simulation Study of ARIMA and Fuzzy Time Series Model for Forecasting Time Series Data

International Journal of Scientific Research in Science Engineering and Technology ◽

10.32628/ijsrset1184112 ◽

2018 ◽

pp. 49-56

Author(s):

Haji A. Haji ◽

Kusman Sadik ◽

Agus Mohamad Soleh

Keyword(s):

Time Series ◽

Simulation Study ◽

Time Series Data ◽

Mean Squared Error ◽

Time Series Model ◽

Fuzzy Time Series ◽

Series Data ◽

Percentage Error ◽

Data Generation ◽

Data Set

Simulation study is used when real world data is hard to ﬁnd or time consuming to gather and it involves generating data set by specific statistical model or using random sampling. A simulation of the process is useful to test theories and understand behavior of the statistical methods. This study aimed to compare ARIMA and Fuzzy Time Series (FTS) model in order to identify the best model for forecasting time series data based on 100 replicates on 100 generated data of the ARIMA (1,0,1) model.There are 16 scenarios used in this study as a combination between 4 data generation variance error values (0.5, 1, 3,5) with 4 ARMA(1,1) parameter values. Furthermore, The performances were evaluated based on three metric mean absolute percentage error (MAPE),Root mean squared error (RMSE) and Bias statistics criterion to determine the more appropriate method and performance of model. The results of the study show a lowest bias for the chen fuzzy time series model and the performance of all measurements is small then other models. The results also proved that chen method is compatible with the advanced forecasting techniques in all of the consided situation in providing better forecasting accuracy.

Download Full-text