Real-time Prediction System Research Articles

Realtime Tsunami Prediction System Using Ocean Floor Network for Local Regions

The ocean floor network system for earthquakes and tsunamis is one of the effective tools for the early detection of large earthquakes on plate boundaries and the tsunamis they generate. The Dense Oceanfloor Network system for Earthquakes and Tsunamis (DONET) was installed in the first rupture areas of the 1944 Tonankai and 1946 Nankai earthquakes. The DONET around the Nankai Trough, a site of huge earthquakes that have caused severe damage, has the potential to detect the genesis of a tsunami. We developed a real-time tsunami prediction system for local communities that takes advantage of the features of DONET, and we have already made it available to several local governments and a commercial company. The outputs of the prediction are the tsunami arrival time, its height, its inundation area, and inundation depth. The system makes real-time monitoring of tsunamis possible. The system should be conceptually applicable to the Nankai Trough area, which has characteristics consistent with the assumptions the system makes about tsunami propagation, crustal activities, and coastal communities. Here, we describe the conceptual basis of the system, the features used to ensure the accuracy of predictions, and the policies used to develop and implement them.

Real-time miscarriage prediction: A comprehensive real-world dataset and a new model

Nowadays, sensors and mobile phones remain important tools for gathering real-time data about human's behaviors and context information; in order to make predictions and advanced preventions of outcomes and diseases. In the specific field of healthcare, real-time decisions are in need to save lives and to make people's lives easier and healthier. In the present paper, we propose a new e-monitoring system for real-time miscarriage prediction that use real time risk factors of miscarriage prediction; which are collected from both healthcare sensors and mobile phones. The challenge of this study is to gather real-time risk factors to make predictions and to propose a model for real time decisions; instead of using only echography because it is often too late to save baby's life. We used two categories of data: data coming from wearables (collected using IoT technologies) and data collected from mobile phones. A mobile phone application that we created collects all risk factors (every 60 s) for analysis and process. Data generated contains real-world data of real pregnant woman for testing and validating the proposed model in term of efficiency and effectiveness. Our dataset of 15 features that represent risk factors of miscarriage, presents an important source that can be used in future researches for further applications. The proposed dataset is available in Mendeley repository through the following link: http://dx.doi.org/10.17632/5sbmhh6t3r.1.

Application of Cloudburst Prediction System: A Review

Abrupt and powerful downpours known as "cloudbursts" can cause disastrous flash floods, landslides, and damage to infrastructure. Twenty-three army troops were washed away and five people killed as a result of the recent flood in Sikkim triggered by a cloudburst over Lhonak Lake in north Sikkim, which overflowed and raised the water levels in the Teesta River. Researchers have created a system based on Convolutional Neural Networks (CNN), weather forecasting, and data mining approaches for weather prediction by modeling meteorological data in order to address the urgent challenge of anticipating and managing cloudburst incidents. The suggested approach in this research describes a real-time cloudburst prediction system that makes use of sophisticated anomaly detection techniques and is based on a machine learning model.

Intelligent Decision Support System Based on Video Recognition of Tuyere Hearth in a Blast Furnace

The paper presents an approach to the development of an intelligent system for real-time prediction of blast-furnace process states. The forecast is based on the analysis of image video sequence obtained as a result of stream video cameras installed on the tuyeres of a blast furnace. Algorithms for recognition of video images of tuyere hearths, as well as scenario forecasting of the evolution of technological situations, are proposed, The research aimes at the analysis of blast-furnace production situations and the prediction of their evolution in real time. This enables timely decisions on adjusting the control in automatic or automated mode. Based on the revealed patterns in the change of video data and by means of the algorithm developed by the authors for process dynamics analysis and prediction, a method is proposed, for early detection of abnormal situations at tuyeres, including the ones enatailing blast furnace destabilization process. Not only the process state at the next time instant is predicted, but also the most probable chain of several subsequent states. For the associative pattern search, Markov chains, machine learning and wavelet analysis methods are used.

Development of a Cloud-Based Real-Time Building Health Monitoring and Prediction System Using AI and IoT Sensors

In this study, a cloud-based real-time building health monitoring and prediction system using AI and IoT sensors was developed. To predict the building condition, which constitutes time-series data, statistical-based ARIMA and AI-based LSTM prediction models were designed, and the effectiveness of the proposed prediction models was experimentally verified using a 1/8-scaled miniaturized structure. The prediction accuracy in terms of MAPE (less than 1%) was experimentally confirmed to be satisfactory. Moreover, a method for analyzing dimensional structure deformation was developed by combining multiple sensor measurements, and its effectiveness was verified through the case study of a real earthquake-damaged building.

Explorations of A Real-Time VR Emotion Prediction System Using Wearable Brain-Computer Interfacing

The following research describes the potential of using a four-class emotion classification using a four-channel wearable EEG headset combined with VR for evoking emotions from each individual. Multiple researchers have conducted and established emotion recognition by using a 2-D monitor screen for stimulus responses but this introduces artifacts such as the lack of concentration on-screen or external noise disturbance and the bulky and cumbersome wires on an EEG device were difficult and time-consuming to set up thus restricting to only the trained professionals to operate this complex and sensitive medical equipment. Therefore, using a small and portable EEG headset where it was accessible for consumers was used for the brainwave signal collection. The wearable EEG headset collects the brainwave samples at 256Hz at specific locations of the brain (Tp9, Tp10, AF7, AF8) and samples were transformed via FFT to obtain the five bands (Delta, Theta, Alpha, Beta, Gamma) and were classified using random forest classifier. An emotion prediction system was then developed and the trained model was used to benchmark the 4-class emotion prediction accuracy from each individual using a 4-channel low-cost EEG headset. Subsequently, a real-time prediction system was implemented and tested. Early findings showed that it could achieve predictions as high as 76.50% for intra-subject classification results.

Toward a real‐time Smart Parking Data Management and Prediction (SPDMP) system by attributes representation learning

Managing and estimating the availability information of parking lots is of great importance to travelers and managers. However, the task is very challenging since the occupancy rate is affected by various factors, including spatial-temporal features, parking lot attributes features, and environmental changes. Previous studies mostly focus on the short-term prediction by capturing the historical sequential dependencies among inputs and outputs, which leads to low estimation accuracy for long-term prediction and limited scalability for real deployment in parking lots. To address the challenges, a comprehensive framework for real-time Smart Parking Data Management and Prediction (SPDMP) system is proposed. Three types of data sources, including historical sequential data, real-time sequential data, and attributes category data, are sufficiently integrated into a customized Parking Availability Prediction (PAP) neural network by representation learning and heterogeneous feature embedding. Specifically, instead of using parking lot property information and environmental data directly, the authors design an Attributes Tensor Embedding Component (ATEC) to integrate the intra-class affection and interclass representations and correlations by two steps of customized embedding process. To balance the impact of the indiscriminate features for various prediction targets, this study proposes a multi-factor attention mechanism to learn the weights and help the PAP achieve a stable performance for time-various tasks. From extensive experiments on two real-world large-scale data sets collected in China and USA (including 19 urban parking and 2 truck parking lots), PAP achieves superior prediction results on both urban parking (6.69% average MAPE) and truck parking prediction (7.63% average MAPE) for five prediction time slots (5 min, 10 min, 30 min, 1 h, and 2 h). Furthermore, even with limited training data, PAP still shows better transferability and adaptability for various types of lots. The proposed SPDMP is selected and used as a pilot test bed by the Washington State Department of Transportation (WSDOT) for truck parking information promotion.

Propagation of a Meteotsunami from the Yellow Sea to the Korea Strait in April 2019

A meteotsunami with a wave height of 0.1–0.9 m and a period of 60 min was observed at tide gauges along the Korea Strait on 7 April 2019, while a train of two to four atmospheric pressure disturbances with disturbance heights of 1.5–3.9 hPa moved eastward from the Yellow Sea to the Korea Strait. Analysis of observational data indicated that isobar lines of the atmospheric pressure disturbances had angles of 75–83° counterclockwise due east and propagated with a velocity of 26.5–31.0 m/s. The generation and propagation process of the meteotsunami was investigated using the Regional Ocean Modeling System. The long ocean waves were amplified due to Proudman resonance in the southwestern Yellow Sea, where the water is deeper than 75 m; here, the long ocean waves were refracted toward the coast on the shallow coastal region of the northern Korea Strait. Refraction and reflection by offshore islands significantly affect the wave heights at the coast. To investigate the effects of an eastward-moving velocity and angle of atmospheric pressure disturbance on the height of a long ocean wave, sensitivity simulations were performed. This result will be useful for the real-time prediction system of meteotsunamis in the Korea Strait.

Real-time forecasting of river water level in urban based on radar rainfall: A case study in Fuzhou City

Urban river network plays an important role in the process of drainage. The real-time prediction system of river water level can provide early warning information to the decision-makers of relevant departments in time, which plays an important role in reducing flood risk. Aiming at the rapid and strong response of urban river to rainfall, a real-time rolling prediction system of urban river water level is proposed based on radar rainfall prediction products and coupling model of hydrology and hydrodynamics. In the system, each simulation is divided into two parts: the review period and the prediction period. In the review period, the river model is driven by the observed rainfall to obtain the latest simulation state to avoid cumulative error, while in the prediction period, the radar rainfall is used to drive in the short-term prediction of river water level. Taking Fuzhou as the research area, a long series review of the process from April 1, 2020 to October 1, 2020 was carried out. The accuracy of the river model, radar-rainfall and prediction system during the study period are evaluated and discussed in detail. The results show that the river model can simulate the process of water level change in rainfall and rain free period. The performance of radar rainfall to total rainfall prediction is better than that of process prediction. The real-time prediction system can give the water level risk according to each radar forecast rainfall process, and has good performance in judging the peak water level and peak time, which can provide reference for urban river dispatching decision-makers.

A Bayesian approach to real-time spatiotemporal prediction systems for bronchiolitis

Respiratory Syncytial Virus (RSV) induced bronchiolitis is a common lung infection and a major cause of infant hospitalization and mortality. Unfortunately, there is no known cure for RSV but several vaccines are in various stages of clinical trials. Currently, immunoprophylaxis is a preventative measure consisting of a series of monthly shots that should be administered at the start, and throughout, peak RSV season. Thus, the successful implementation of immunoprophylaxis is contingent upon understanding when outbreak seasons will begin, peak, and end. In this research we estimate the seasonal epidemic curves of RSV induced bronchiolitis using a spatially varying change point model. Further, in a novel approach and using the fitted change point model, we develop a historical matching algorithm to generate real time predictions of seasonal curves for future years.

Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal.

The aim of the study was to inspect the acoustic properties and sleep characteristics of a preapneic snoring sound. The feasibility of forecasting upcoming respiratory events by snoring sound was also investigated. Participants with habitual snoring or a heavy breathing sound during sleep were recruited consecutively. Polysomnography was conducted, and snoring-related breathing sound was recorded simultaneously. Acoustic features and sleep features were extracted from 30-second samples, and a machine learning algorithm was used to establish 2 prediction models. A total of 74 eligible participants were included. Model 1, tested by 5-fold cross-validation, achieved an accuracy of 0.92 and an area under the curve of 0.94 for respiratory event prediction. Model 2, with acoustic features and sleep information tested by Leave-One-Out cross-validation, had an accuracy of 0.78 and an area under the curve of 0.80. Sleep position was found to be the most important among all sleep features contributing to the performance of the 2 models. Preapneic sound presented unique acoustic characteristics, and snoring-related breathing sound could be deployed as a real-time apneic event predictor. The models, combined with sleep information, serve as a promising tool for an early warning system to forecast apneic events. Wang B, Yi X, Gao J, etal. Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal. J Clin Sleep Med. 2021;17(9):1777-1784.

Development Trend of Intelligent Transportation in the Era of Big Data

The arrival of the big data era has brought great convenience to people’s real life. All aspects of our lives follow the development of the big data era. In the context of the era of big data, through data collection and analysis, the current urban transportation has been reasonably developed, laying a good foundation for the development of intelligent transportation in the future, and can solve various current problems. Based on the analysis of the traffic system, this article finds that the real-time road condition prediction of common functions is highly correlated with route guidance. Therefore, this paper uses a method based on historical data to predict road conditions, and uses the prediction results to dynamically allocate road segments, and then uses the relevant single-source shortest path algorithm to simulate the optimal path. By analyzing the application advantages of data mining in intelligent transportation, Baum Welch’s algorithm is used for data mining in a real-time traffic prediction system based on historical data. Real-time road conditions are matched with historical data models to predict road conditions. Finally, analyzed and designed the big data cloud platform framework applied to it to achieve higher efficiency. The experimental research results show that, by studying the development of urban intelligent transportation systems, the purpose of this paper is to make better use of existing resources to process data more efficiently, and then combine distributed data mining algorithms to optimize the processed data and apply it to in the intelligent transportation system, an intelligent management and control system that is more efficient, accurate, and user experience better can be established.

Real-Time System Prediction for Heart Rate Using Deep Learning and Stream Processing Platforms

Low heart rate causes a risk of death, heart disease, and cardiovascular diseases. Therefore, monitoring the heart rate is critical because of the heart’s function to discover its irregularity to detect the health problems early. Rapid technological advancement (e.g., artificial intelligence and stream processing technologies) allows healthcare sectors to consolidate and analyze massive health-based data to discover risks by making more accurate predictions. Therefore, this work proposes a real-time prediction system for heart rate, which helps the medical care providers and patients avoid heart rate risk in real time. The proposed system consists of two phases, namely, an offline phase and an online phase. The offline phase targets developing the model using different forecasting techniques to find the lowest root mean square error. The heart rate time-series dataset is extracted from Medical Information Mart for Intensive Care (MIMIC-II). Recurrent neural network (RNN), long short-term memory (LSTM), gated recurrent units (GRU), and bidirectional long short-term memory (BI-LSTM) are applied to heart rate time series. For the online phase, Apache Kafka and Apache Spark have been used to predict the heart rate in advance based on the best developed model. According to the experimental results, the GRU with three layers has recorded the best performance. Consequently, GRU with three layers has been used to predict heart rate 5 minutes in advance.

Neuro-feedback system for real-time BCI decision prediction

Neuro-feedback system for real-time BCI decision prediction

Real-time prediction of interstitial oxygen concentration in Czochralski silicon using machine learning

We developed a machine learning model to predict interstitial oxygen (Oi) concentration in a Czochralski-grown silicon crystal. A highly accurate prediction can be ensured by selecting the appropriate experimental parameters that represent the change in the furnace conditions. A neural network was trained using the dataset of 450 ingots, and its prediction error for the testing dataset was 4.2 × 1016 atoms cm−3. Finally, a real-time prediction system was developed wherein the crystal growth data are input into the model, and the Oi concentration at the current growth interface is calculated immediately.

Human Falling Detection Algorithm Based on Multisensor Data Fusion with SVM

Falling is a common phenomenon in the life of the elderly, and it is also one of the 10 main causes of serious health injuries and death of the elderly. In order to prevent falling of the elderly, a real-time fall prediction system is installed on the wearable intelligent device, which can timely trigger the alarm and reduce the accidental injury caused by falls. At present, most algorithms based on single-sensor data cannot accurately describe the fall state, while the fall detection algorithm based on multisensor data integration can improve the sensitivity and specificity of prediction. In this study, we design a fall detection system based on multisensor data fusion and analyze the four stages of falls using the data of 100 volunteers simulating falls and daily activities. In this paper, data fusion method is used to extract three characteristic parameters representing human body acceleration and posture change, and the effectiveness of the multisensor data fusion algorithm is verified. The sensitivity is 96.67%, and the specificity is 97%. It is found that the recognition rate is the highest when the training set contains the largest number of samples in the training set. Therefore, after training the model based on a large amount of effective data, its recognition ability can be improved, and the prevention of fall possibility will gradually increase. In order to compare the applicability of random forest and support vector machine (SVM) in the development of wearable intelligent devices, two fall posture recognition models were established, respectively, and the training time and recognition time of the models are compared. The results show that SVM is more suitable for the development of wearable intelligent devices.

Prediction of spatiotemporal stability and rainfall threshold of shallow landslides using the TRIGRS and Scoops3D models

The stability evaluation of rainfall-induced landslides using a physical determination model supports disaster prevention, but it is mostly applied to the area with few landslides, and there is a lack of quantitative study on rainfall and landslide stability. This paper combined the Scoops3D model with the TRIGRS model (3D) to predict the shallow landslide spatiotemporal distribution and compared the simulation results with those of the TRIGRS model alone (1D), aiming to obtain more accurate assessment results. At the same time, the relationship between landslide stability and accumulative rainfall was quantitatively fitted to improve the real-time landslide prediction system. We applied the 1D and 3D models to the July 21, 2013 group-occurring landslide event (976 shallow landslides) in the Niangniangba basin, China. The required geotechnical parameters of both models were acquired by field and laboratory tests. We calculated the pressure head over time using the TRIGRS model based on practical rainfall data and predicted the shallow landslide stability using the Scoops3D model according to the resulting piezometric surface. We compared the landslide stability spatial distributions of the two models under initial and saturated conditions with the landslide catalogue. The success rate of landslides predicted by 3D model is 4.20% higher than 1D model. A composite index to quantitatively evaluate both models’ performances indicates over-prediction by the 1D model in the stable region, while the 3D model more effectively predicts shallow landslides with a smaller unstable region. The relationship between instability proportion and accumulative rainfall in the 1D and 3D model can be represented by y=24.57x0.18 and y=11.59x0.33, respectively. The 3D model shows more conservative result, and the rainfall threshold analysis proposed in this paper can provide reference for the time of most landslides in the case of insufficient data, which is an important indicator for disaster early warning and prediction.

Freezing of gait detection in Parkinson's disease via multimodal analysis of EEG and accelerometer signals.

Parkinson's disease (PD) patients with freezing of gait (FOG) can suddenly lose their forward moving ability leading to unexpected falls. To overcome FOG and avoid the falls, a real-time accurate FOG detection or prediction system is desirable to trigger on-demand cues. In this study, we designed and implemented an in-place movement experiment for PD patients to provoke FOG and meanwhile acquired multimodal physiological signals, such as electroencephalography (EEG) and accelerometer signals. A multimodal model using brain activity from EEG and motion data from accelerometers was developed to improve FOG detection performance. In the detection of over 700 FOG episodes observed in the experiments, the multimodal model achieved 0.211 measured by Matthews Correlation Coefficient (MCC) compared with the single-modal models (0.127 or 0.139).Clinical Relevance- This is the first study to use multimodal: EEG and accelerometer signal analysis in FOG detection, and an improvement was achieved.

Real-Time Prediction of Docker Container Resource Load Based on a Hybrid Model of ARIMA and Triple Exponential Smoothing

More and more enterprises are beginning to use Docker containers to build cloud platforms. Predicting the resource usage of container workload has been an important and challenging problem to improve the performance of cloud computing platform. The existing prediction models either incur large time overhead or have insufficient accuracy. This article proposes a hybrid model of the ARIMA and triple exponential smoothing. It can accurately predict both linear and nonlinear relationships in the container resource load sequence. To deal with the dynamic Docker container resource load, the weighting values of the two single models in the hybrid model are chosen according to the sum of squares of their predicted errors for a period of time. We also design and implement a real-time prediction system that consists of the collection, storage, prediction of Docker container resource load data and scheduling optimization of CPU and memory resource usage based on predicted values. The experimental results show that the predicting accuracy of the hybrid model improves by 52.64, 20.15, and 203.72 percent on average compared to the ARIMA, the triple exponential smoothing model and ANN+SaDE model respectively with a small time overhead.

A new Internet of Things architecture for real-time prediction of various diseases using machine learning on big data environment

A number of technologies enabled by Internet of Thing (IoT) have been used for the prevention of various chronic diseases, continuous and real-time tracking system is a particularly important one. Wearable medical devices with sensor, health cloud and mobile applications have continuously generating a huge amount of data which is often called as streaming big data. Due to the higher speed of the data generation, it is difficult to collect, process and analyze such massive data in real-time in order to perform real-time actions in case of emergencies and extracting hidden value. using traditional methods which are limited and time-consuming. Therefore, there is a significant need to real-time big data stream processing to ensure an effective and scalable solution. In order to overcome this issue, this work proposes a new architecture for real-time health status prediction and analytics system using big data technologies. The system focus on applying distributed machine learning model on streaming health data events ingested to Spark streaming through Kafka topics. Firstly, we transform the standard decision tree (DT) (C4.5) algorithm into a parallel, distributed, scalable and fast DT using Spark instead of Hadoop MapReduce which becomes limited for real-time computing. Secondly, this model is applied to streaming data coming from distributed sources of various diseases to predict health status. Based on several input attributes, the system predicts health status, send an alert message to care providers and store the details in a distributed database to perform health data analytics and stream reporting. We measure the performance of Spark DT against traditional machine learning tools including Weka. Finally, performance evaluation parameters such as throughput and execution time are calculated to show the effectiveness of the proposed architecture. The experimental results show that the proposed system is able to effectively process and predict real-time and massive amount of medical data enabled by IoT from distributed and various diseases.