scholarly journals Analyzing Sensor-Based Individual and Population Behavior Patterns via Inverse Reinforcement Learning

Sensors ◽  
2020 ◽  
Vol 20 (18) ◽  
pp. 5207
Author(s):  
Beiyu Lin ◽  
Diane J. Cook
Keyword(s):  
Time Series ◽  
Reinforcement Learning ◽  
Smart Home ◽  
Time Series Data ◽  
Population Level ◽  
Sensor Data ◽  
Series Data ◽  
Inverse Reinforcement Learning ◽  
Daily Routines ◽  
Health Diagnosis

Digital markers of behavior can be continuously created, in everyday settings, using time series data collected by ambient sensors. The goal of this work was to perform individual- and population-level behavior analysis from such time series sensor data. In this paper, we introduce a novel algorithm—Resident Relative Entropy-Inverse Reinforcement Learning (RRE-IRL)—to perform an analysis of a single smart home resident or a group of residents, using inverse reinforcement learning. By employing this method, we learnt an individual’s behavioral routine preferences. We then analyzed daily routines for an individual and for eight smart home residents grouped by health diagnoses. We observed that the behavioral routine preferences changed over time. Specifically, the probability that the observed behavior was the same at the beginning of data collection as it was at the end (months later) was lower for residents experiencing cognitive decline than for cognitively healthy residents. When comparing aggregated behavior between groups of residents from the two diagnosis groups, the behavioral difference was even greater. Furthermore, the behavior preferences were used by a random forest classifier to predict a resident’s cognitive health diagnosis, with an accuracy of 0.84.

scholarly journals Identification of Animal Behavioral Strategies by Inverse Reinforcement Learning

2017 ◽  
Author(s):  
Shoichiro Yamaguchi ◽  
Honda Naoki ◽  
Muneki Ikeda ◽  
Yuki Tsukada ◽  
Shunji Nakano ◽  
...  
Keyword(s):  
Decision Making ◽  
Time Series ◽  
Reinforcement Learning ◽  
Time Series Data ◽  
Series Data ◽  
Behavioral Strategy ◽  
Behavioral Strategies ◽  
Inverse Reinforcement Learning ◽  
C Elegans ◽  
Temporal Derivative

AbstractAnimals are able to reach a desired state in an environment by controlling various behavioral patterns. Identification of the behavioral strategy used for this control is important for understanding animals’ decision-making and is fundamental to dissect information processing done by the nervous system. However, methods for quantifying such behavioral strategies have not been fully established. In this study, we developed an inverse reinforcement-learning (IRL) framework to identify an animal’s behavioral strategy from behavioral time-series data. As a particular target, we applied this framework to C. elegans thermotactic behavior; after cultivation at a constant temperature with or without food, the fed and starved worms prefer and avoid from the cultivation temperature on a thermal gradient, respectively. Our IRL approach revealed that the fed worms used both absolute and temporal derivative of temperature and that their strategy comprised mixture of two strategies: directed migration (DM) and isothermal migration (IM). The DM is a strategy that the worms efficiently reach to specific temperature, which explained thermotactic behaviors of the fed worms. The IM is a strategy that the worms track along a constant temperature, which reflects isothermal tracking well observed in previous studies. We also showed the neural basis underlying the strategies, by applying our method to thermosensory neuron-deficient worms. In contrast to fed animals, the strategy of starved animals indicated that they escaped the cultivation temperature using only absolute, but not temporal derivative of temperature. Thus, our IRL-based approach is capable of identifying animal strategies from behavioral time-series data and will be applicable to wide range of behavioral studies, including decision-making of other organisms.Author SummaryUnderstanding animal decision-making has been a fundamental problem in neuroscience and behavioral ecology. Many studies analyze actions that represent decision-making in behavioral tasks, in which rewards are artificially designed with specific objectives. However, it is impossible to extend this artificially designed experiment to a natural environment, because in a natural environment, the rewards for freely-behaving animals cannot be clearly defined. To this end, we must reverse the current paradigm so that rewards are identified from behavioral data. Here, we propose a new reverse-engineering approach (inverse reinforcement learning) that can estimate a behavioral strategy from time-series data of freely-behaving animals. By applying this technique with thermotaxis in C. elegans, we successfully identified the reward-based behavioral strategy.

scholarly journals Remaining Useful Life Prediction Using Temporal Convolution with Attention

AI ◽  
2021 ◽  
Vol 2 (1) ◽  
pp. 48-70
Author(s):  
Wei Ming Tan ◽  
T. Hui Teo
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Time Series Data ◽  
Remaining Useful Life ◽  
Sensor Data ◽  
Series Data ◽  
Multiple Time ◽  
Data Set ◽  
Form Complex ◽  
Useful Life

Prognostic techniques attempt to predict the Remaining Useful Life (RUL) of a subsystem or a component. Such techniques often use sensor data which are periodically measured and recorded into a time series data set. Such multivariate data sets form complex and non-linear inter-dependencies through recorded time steps and between sensors. Many current existing algorithms for prognostic purposes starts to explore Deep Neural Network (DNN) and its effectiveness in the field. Although Deep Learning (DL) techniques outperform the traditional prognostic algorithms, the networks are generally complex to deploy or train. This paper proposes a Multi-variable Time Series (MTS) focused approach to prognostics that implements a lightweight Convolutional Neural Network (CNN) with attention mechanism. The convolution filters work to extract the abstract temporal patterns from the multiple time series, while the attention mechanisms review the information across the time axis and select the relevant information. The results suggest that the proposed method not only produces a superior accuracy of RUL estimation but it also trains many folds faster than the reported works. The superiority of deploying the network is also demonstrated on a lightweight hardware platform by not just being much compact, but also more efficient for the resource restricted environment.

A Novel Approach to Time Series Forecasting Using Model-Free Adaptive Control Framework

2020 ◽  
Author(s):  
Meenakshi Narayan ◽  
Ann Majewicz Fey
Keyword(s):  
Time Series ◽  
Adaptive Control ◽  
Real Time ◽  
Time Series Data ◽  
Mean Squared Error ◽  
Force Sensor ◽  
Sensor Data ◽  
Series Data ◽  
Model Free ◽  
Control Framework

Abstract Sensor data predictions could significantly improve the accuracy and effectiveness of modern control systems; however, existing machine learning and advanced statistical techniques to forecast time series data require significant computational resources which is not ideal for real-time applications. In this paper, we propose a novel forecasting technique called Compact Form Dynamic Linearization Model-Free Prediction (CFDL-MFP) which is derived from the existing model-free adaptive control framework. This approach enables near real-time forecasts of seconds-worth of time-series data due to its basis as an optimal control problem. The performance of the CFDL-MFP algorithm was evaluated using four real datasets including: force sensor readings from surgical needle, ECG measurements for heart rate, and atmospheric temperature and Nile water level recordings. On average, the forecast accuracy of CFDL-MFP was 28% better than the benchmark Autoregressive Integrated Moving Average (ARIMA) algorithm. The maximum computation time of CFDL-MFP was 49.1ms which was 170 times faster than ARIMA. Forecasts were best for deterministic data patterns, such as the ECG data, with a minimum average root mean squared error of (0.2±0.2).

scholarly journals Anonymizing Sensor Data on the Edge: A Representation Learning and Transformation Approach

2022 ◽  
Vol 3 (1) ◽  
pp. 1-26
Author(s):  
Omid Hajihassani ◽  
Omid Ardakanian ◽  
Hamzeh Khazaei
Keyword(s):  
Time Series ◽  
Linear Transformation ◽  
Input Data ◽  
Time Series Data ◽  
Representation Learning ◽  
Sensor Data ◽  
Series Data ◽  
Sensitive Information ◽  
Latent Space ◽  
Low Dimensional

The abundance of data collected by sensors in Internet of Things devices and the success of deep neural networks in uncovering hidden patterns in time series data have led to mounting privacy concerns. This is because private and sensitive information can be potentially learned from sensor data by applications that have access to this data. In this article, we aim to examine the tradeoff between utility and privacy loss by learning low-dimensional representations that are useful for data obfuscation. We propose deterministic and probabilistic transformations in the latent space of a variational autoencoder to synthesize time series data such that intrusive inferences are prevented while desired inferences can still be made with sufficient accuracy. In the deterministic case, we use a linear transformation to move the representation of input data in the latent space such that the reconstructed data is likely to have the same public attribute but a different private attribute than the original input data. In the probabilistic case, we apply the linear transformation to the latent representation of input data with some probability. We compare our technique with autoencoder-based anonymization techniques and additionally show that it can anonymize data in real time on resource-constrained edge devices.

scholarly journals Matrix Profile-Based Approach to Industrial Sensor Data Analysis Inside RDBMS

Mathematics ◽  
2021 ◽  
Vol 9 (17) ◽  
pp. 2146
Author(s):  
Mikhail Zymbler ◽  
Elena Ivanova
Keyword(s):  
Time Series ◽  
Time Series Data ◽  
Wide Spectrum ◽  
Database Management System ◽  
Third Party ◽  
Sensor Data ◽  
Data Conversion ◽  
Series Data ◽  
Time Interval ◽  
Short Time Interval

Currently, big sensor data arise in a wide spectrum of Industry 4.0, Internet of Things, and Smart City applications. In such subject domains, sensors tend to have a high frequency and produce massive time series in a relatively short time interval. The data collected from the sensors are subject to mining in order to make strategic decisions. In the article, we consider the problem of choosing a Time Series Database Management System (TSDBMS) to provide efficient storing and mining of big sensor data. We overview InfluxDB, OpenTSDB, and TimescaleDB, which are among the most popular state-of-the-art TSDBMSs, and represent different categories of such systems, namely native, add-ons over NoSQL systems, and add-ons over relational DBMSs (RDBMSs), respectively. Our overview shows that, at present, TSDBMSs offer a modest built-in toolset to mine big sensor data. This leads to the use of third-party mining systems and unwanted overhead costs due to exporting data outside a TSDBMS, data conversion, and so on. We propose an approach to managing and mining sensor data inside RDBMSs that exploits the Matrix Profile concept. A Matrix Profile is a data structure that annotates a time series through the index of and the distance to the nearest neighbor of each subsequence of the time series and serves as a basis to discover motifs, anomalies, and other time-series data mining primitives. This approach is implemented as a PostgreSQL extension that allows an application programmer both to compute matrix profiles and mining primitives and to represent them as relational tables. Experimental case studies show that our approach surpasses the above-mentioned out-of-TSDBMS competitors in terms of performance since it assumes that sensor data are mined inside a TSDBMS at no significant overhead costs.

scholarly journals Simple IoT simulation

2019 ◽  
Vol 9 (1) ◽  
pp. 3328-3330
Keyword(s):  
Time Series ◽  
Large Scale ◽  
Time Series Data ◽  
Random Number Generator ◽  
Power Supplies ◽  
Sensor Data ◽  
Series Data ◽  
Scaled Model ◽  
True Time ◽  
Server Architecture

While analyzing iot projects it is very expensive to buy a lot of sensors , corresponding processor boards, power supplies etc. Moreover the entire process is to be replicated to cater to large topologies. The whole experiment is to be planned at a large scale before we can actually start to see analytics working. At a smaller scale this can be implemented as a simulation program in linux where the sensor data is created using a random number generator and scaled appropriately for each type of sensor to mimic representative data. This is them encrypted before sending it over the network to the edge nodes. At the server a socket stream now continuously awaits sensor data Here the required sensor data is retrieved and decrypted to give true time series data. This time series is now given to an analytics engine which calculates the trends and cyclicity and is used to train a neural network. The anomalies so found are properly deciphered. The multiplicity of the nodes can be characterized by having several client programs running in separate terminals. A simple client server architecture is thus able to simulate a large iot infrastructure and is able to perform analytics on a scaled model

scholarly journals Challenges in estimating the impact of vaccination with sparse data

2018 ◽  
Author(s):  
Kayoko Shioda ◽  
Cynthia Schuck-Paim ◽  
Robert J. Taylor ◽  
Roger Lustig ◽  
Lone Simonsen ◽  
...  
Keyword(s):  
Time Series ◽  
Time Series Data ◽  
Simulation Analysis ◽  
Principal Component ◽  
Population Level ◽  
Series Data ◽  
Synthetic Control ◽  
Vaccine Impact ◽  
Pca Method ◽  
The Impact

ABSTRACTBackgroundThe synthetic control (SC) model is a powerful tool to quantify the population-level impact of vaccines, because it can adjust for trends unrelated to vaccination using a composite of control diseases. Because vaccine impact studies are often conducted using smaller subnational datasets, we evaluated the performance of SC models with sparse time series data. To obtain more robust estimates of vaccine effects from noisy time series, we proposed a possible alternative approach, “STL+PCA” method (seasonal-trend decomposition plus principal component analysis), which first extracts smoothed trends from the control time series and uses them to adjust the outcome.MethodsUsing both the SC and STL+PCA models, we estimated the impact of 10-valent pneumococcal conjugate vaccine (PCV10) on pneumonia hospitalizations among cases <12 months and 80+ years of age during 2004-2014 at the subnational level in Brazil. The performance of these models was also compared using simulation analyses.ResultsThe SC model was able to adjust for trends unrelated to PCV10 in larger states but not in smaller states. The simulation analysis confirmed that the SC model failed to select an appropriate set of control diseases when the time series were sparse and noisy, thereby generating biased estimates of the impact of vaccination when secular trends were present. The STL+PCA approach decreased bias in the estimates for smaller populations.ConclusionsEstimates from the SC model might be biased when data are sparse. The STL+PCA model provides more accurate evaluations of vaccine impact in smaller populations.

scholarly journals Remaining Useful Life Estimation from Event Data

2021 ◽  
Vol 13 (1) ◽  
Author(s):  
Mahbubul Alam ◽  
Laleh Jalali ◽  
Mahbubul Alam ◽  
Ahmed Farahat ◽  
Chetan Gupta
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Time Series Data ◽  
Remaining Useful Life ◽  
Sensor Data ◽  
Series Data ◽  
Sequential Data ◽  
Event Data ◽  
Useful Life ◽  
Event Based

Abstract—Prognostics aims to predict the degradation of equipment by estimating their remaining useful life (RUL) and/or the failure probability within a specific time horizon. The high demand of equipment prognostics in the industry have propelled researchers to develop robust and efficient prognostics techniques. Among data driven techniques for prognostics, machine learning and deep learning (DL) based techniques, particularly Recurrent Neural Networks (RNNs) have gained significant attention due to their ability of effectively representing the degradation progress by employing dynamic temporal behaviors. RNNs are well known for handling sequential data, especially continuous time series sequential data where the data follows certain pattern. Such data is usually obtained from sensors attached to the equipment. However, in many scenarios sensor data is not readily available and often very tedious to acquire. Conversely, event data is more common and can easily be obtained from the error logs saved by the equipment and transmitted to a backend for further processing. Nevertheless, performing prognostics using event data is substantially more difficult than that of the sensor data due to the unique nature of event data. Though event data is sequential, it differs from other seminal sequential data such as time series and natural language in the following manner, i) unlike time series data, events may appear at any time, i.e., the appearance of events lacks periodicity; ii) unlike natural languages, event data do not follow any specific linguistic rule. Additionally, there may be a significant variability in the event types appearing within the same sequence.  Therefore, this paper proposes an RUL estimation framework to effectively handle the intricate and novel event data. The proposed framework takes discrete events generated by an equipment (e.g., type, time, etc.) as input, and generates for each new event an estimate of the remaining operating cycles in the life of a given component. To evaluate the efficacy of our proposed method, we conduct extensive experiments using benchmark datasets such as the CMAPSS data after converting the time-series data in these datasets to sequential event data. The event data conversion is carried out by careful exploration and application of appropriate transformation techniques to the time series. To the best of our knowledge this is the first time such event-based RUL estimation problem is introduced to the community. Furthermore, we propose several deep learning and machine learning based solution for the event-based RUL estimation problem. Our results suggest that the deep learning models, 1D-CNN, LSTM, and multi-head attention show similar RMSE, MAE and Score performance. Foreseeably, the XGBoost model achieve lower performance compared to the deep learning models since the XGBoost model fails to capture ordering information from the sequence of events. 

Sign in / Sign up

Export Citation Format

Share Document