Anomaly Detection on Data Streams for Smart Agriculture

Moso, Juliet Chebet; Cormier, Stéphane; Runz, Cyril de; Fouchal, Hacène; Wandeto, John M.

doi:10.3390/agriculture11111083

Cited by 30 publications

(16 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Winsorization, thus, is a crucial approach for detecting significant occurrences in the farm data. It is the process of identifying observations that deviate from the norm (Moso et al, 2021).…”

Section: Resultsmentioning

confidence: 99%

Outlier Removal in Sheep Farm Datasets Using Winsorization

Hamadani

Ganai

Raja

et al. 2022

BKAP

View full text Add to dashboard Cite

Background: Sheep farm data is often biased by extreme values which are generally introduced due to errors in manual measurement. These values interfere with the accuracy of estimations especially in state-of-the-art techniques like Machine Learning. Methods: Therefore, winsorization technique was attempted for the removal of outliers from sheep farm data data for 11 years (2011-2021) for body weights at different ages. Some outliers were deliberately introduced into the data to check the efficiency of the technique. This study was conducted during the year 2021. Result: Our results indicate that outlier values of 15.3, 42, 44, 60, 90 for birth weight, weaning weight, 6-month, 9 month and 12-month body weight which were far from the normal range were removed using this technique. The mean and standard deviation values were altered after winsorization. Winsorization technique works well for sheep farm data to remove the bias introduced by outliers and also removes, to a large extent, the need for manual outlier removal in data.

show abstract

Section: Resultsmentioning

confidence: 99%

Outlier Removal in Sheep Farm Datasets Using Winsorization

Hamadani

Ganai

Raja

et al. 2022

BKAP

View full text Add to dashboard Cite

show abstract

“…These generated data must be protected from adversarial attacks to enhance agricultural productivity. Moso et al [23] introduced an ensemble anomaly detector called Enhanced Local Selective Combination in Parallel Outlier Ensembles (ELSCP) to accomplish the task. A data-driven unsupervised methodology was presented that was applied to two different case studies, where one dealt with global positioning system (GPS) traces and the other dealt with crop data.…”

Section: Related Workmentioning

confidence: 99%

Intrusion Detection in Internet of Things Based Smart Farming Using Hybrid Deep Learning Framework

Kethineni

Pradeepini

2023

Preprint

View full text Add to dashboard Cite

Smart agriculture is a popular domain due to its intensified growth in recent times. This domain aggregates the advantages of several computing technologies, where the IoT is the most popular and beneficial. In this work, a novel and effective deep learning based framework is developed to detect intrusions in smart farming systems. The architecture is three-tier, with the first tier being the sensor layer, which involves the placement of sensors in agricultural areas. The second tier is the Fog Computing Layer (FCL), which consists of Fog nodes, and the proposed IDS is implemented in each Fog node. The gathered information is transferred to this fog layer for further analysis of data. The third tier is the cloud computing layer, which provides data storage and end-to-end services. The proposed model includes a fused CNN model with the bidirectional gated recurrent unit (Bi-GRU) model to detect and classify intruders. An attention mechanism is included within the BiGRU model to find the key features responsible for identifying the DDoS attack. In addition, the accuracy of the classification model is improved by using a nature-inspired meta-heuristic optimization algorithm called the Wild Horse Optimization (WHO) algorithm. The last layer is the cloud layer, which collects data from fog nodes and offers storage services. The proposed system will be implemented in the Python platform, using ToN-IoT and APA-DDoS attack datasets for assessment. The proposed system outperforms the existing methods in accuracy (99.35%), detection rate (98.99%), precision (99.9%) and F-Score (99.08%) for the APA DDoS attack dataset and the achieved accuracy of the ToN-IoT dataset (99.71%), detection rate (99.02%), precision (99.89%) and F-score (99.05%).

show abstract

“…Anomaly detection for time series data can also be applied in crop harvesting. Moso et al (2021) proposed a powerful ensemble-based approach for anomaly detection, which was mainly used for data streams generated in smart agriculture. This technology can be applied to crop data sets and identify anomalies that affect crop harvest.…”

Section: Related Workmentioning

confidence: 99%

“…This can also expose the network to attacks that could lead to data tampering ( Abdallah et al, 2021 ). Missing or misrepresented data is significantly different from normal data in the time series data collected by the sensors ( Moso et al, 2021 ). These can be considered as anomalies in the data ( Adkisson et al, 2021 ).…”

Section: Introductionmentioning

confidence: 99%

Anomaly Detection for Internet of Things Time Series Data Using Generative Adversarial Networks With Attention Mechanism in Smart Agriculture

Cheng

Wang

et al. 2022

Front. Plant Sci.

View full text Add to dashboard Cite

More recently, smart agriculture has received widespread attention, which is a deep combination of modern agriculture and the Internet of Things (IoT) technology. To achieve the aim of scientific cultivation and precise control, the agricultural environments are monitored in real time by using various types of sensors. As a result, smart agricultural IoT generated a large amount of multidimensional time series data. However, due to the limitation of applied scenarios, smart agricultural IoT often suffers from data loss and misrepresentation. Moreover, some intelligent decision-makings for agricultural management also require the detailed analysis of data. To address the above problems, this article proposes a new anomaly detection model based on generative adversarial networks (GAN), which can process the multidimensional time series data generated by smart agricultural IoT. GAN is a deep learning model to learn the distribution patterns of normal data and capture the temporal dependence of time series and the potential correlations between features through learning. For the problem of generator inversion, an encoder–decoder structure incorporating the attention mechanism is designed to improve the performance of the model in learning normal data. In addition, we also present a new reconstruction error calculation method that measures the error in terms of both point-wise difference and curve similarity to improve the detection effect. Finally, based on three smart agriculture-related datasets, experimental results show that our proposed model can accurately achieve anomaly detection. The experimental precision, recall, and F1 score exceeded the counterpart models by reaching 0.9351, 0.9625, and 0.9482, respectively.

show abstract

Anomaly Detection on Data Streams for Smart Agriculture

Cited by 30 publications

References 40 publications

Outlier Removal in Sheep Farm Datasets Using Winsorization

Outlier Removal in Sheep Farm Datasets Using Winsorization

Intrusion Detection in Internet of Things Based Smart Farming Using Hybrid Deep Learning Framework

Anomaly Detection for Internet of Things Time Series Data Using Generative Adversarial Networks With Attention Mechanism in Smart Agriculture

Contact Info

Product

Resources

About