Python library used in the elaboration of the paper: "Data Reconstruction Applications for IoT Air Pollution Sensor Networks Using Graph Signal Processing" published in the Journal of Network and Computer Applications (JNCA, Elsevier). The Python library implements the proposed graph-based data reconstruction framework which is based on the graph learning model defined by Dong et al. 2016 and the graph interpolated regularization defined by Belkin et al. 20004. This python library can be applied to sensor networks to carry out different post-processing applications such as; missing value imputation, creation of virtual sensors, data fusion, or drift compensation among others.
Authors: Pau Ferrer-Cid, Jose M. Barcelo-Ordinas and Jorge Garcia-Vidal.
Python library used in the elaboration of the paper: "Volterra Graph-Based Outlier Detection for Air Pollution Sensor Networks" published in the IEEE Transactions on Network Science and Engineering. The Python library implements the proposed Volterra Graph-Based Outlier Detection (VGOD) algorithm along with the Volterra-like graph signal reconstruction model defined by Xiao et al. (2021) and the graph learning optimization problem defined by Dong et al. (2016). Further details on the references are available in the repository.
Authors: Pau Ferrer-Cid, Jose M. Barcelo-Ordinas and Jorge Garcia-Vidal.
Python library used in the elaboration of the paper: "Graph Signal Reconstruction Techniques for Air Pollution Monitoring Platforms" submitted to the IEEE Internet of Things Journal. The Python library implements the different methods used in the paper; laplacian-based interpolated regularization from Belkin et al., GSP low-pass based from Stankovic et al. and graph Kernel Ridge Regression from Romero et al. Finally, the file includes useful functions such as various graph kernels. TO BE PUBLISHED WHEN ACCEPTED.
Authors: Pau Ferrer-Cid, Jose M. Barcelo-Ordinas and Jorge Garcia-Vidal.
Calibration software and data sets used in the elaboration of the paper: "Multi-sensor data fusion calibration in IoT air pollution platforms" submitted to the IEEE Internet of Things journal. There are two types of data sets; the first type contain data from several metal-oxide sensors that coincided in the same location, along with the ozone reference values for calibration purposes. While the second type contain data from one electro-chemical sensor and four metal-oxide sensors, with the ozone reference values.
Authors: Pau Ferrer-Cid, Jose M. Barcelo-Ordinas, Jorge Garcia-Vidal, Anna Ripoll and Mar Viana.
Calibration software and data sets used in the elaboration of the paper: "A comparative study of calibration methods for low-cost ozone sensors in IoT platforms" published in the IEEE Internet of Things journal. Four machine learning models are implemented for the calibration of ozone sensor. The models available are: the Multiple Linear Regression (MLR), the K-Nearest Neighbors (KNN), the Random Forest (RF) and the Support Vector Regression (SVR). The data sets correspond to nodes containing ozone MOX and EC sensors.
Authors: Pau Ferrer-Cid, Jose M. Barcelo-Ordinas, Jorge Garcia-Vidal, Anna Ripoll and Mar Viana.