• Corpus ID: 246275932

A deep mixture density network for outlier-corrected interpolation of crowd-sourced weather data

  title={A deep mixture density network for outlier-corrected interpolation of crowd-sourced weather data},
  author={Charlie Kirkwood and Theodoros Economou and Henry M. Odbert and Nicolas Pugeault},
As the costs of sensors and associated IT infrastructure decreases — as exemplified by the Internet of Things — increasing volumes of observational data are becoming available for use by environmental scientists. However, as the number of available observation sites increases, so too does the opportunity for data quality issues to emerge, particularly given that many of these sensors do not have the benefit of official maintenance teams. To realise the value of crowd sourced ‘Internet of Things… 


Bayesian Deep Learning for Spatial Interpolation in the Presence of Auxiliary Information
The power of feature learning in a geostatistical context is demonstrated by showing how deep neural networks can automatically learn the complex high-order patterns by which point-sampled target variables relate to gridded auxiliary variables and in doing so produce detailed maps.
Development and Application of a Statistically-Based Quality Control for Crowdsourced Air Temperature Data
In this study, a statistically-based quality control (QC) is developed to identify suspicious air temperature (T) measurements from crowdsourced data sets, and the transferability of the newly developed QC and the use of CWS data is illustrated.
Deep covariate-learning: optimising information extraction from terrain texture for geostatistical modelling applications
A deep learning approach to automatically deriving optimal task-specific terrain texture covariates from a standard SRTM 90m gridded digital elevation model (DEM) is presented, finding that this approach produces covariates for geostatistical modelling that have surprisingly strong explanatory power on their own.
Can the crowdsourcing data paradigm take atmospheric science to a new level? A case study of the urban heat island of London quantified using Netatmo weather stations
Crowdsourcing techniques are frequently used across science to supplement traditional means of data collection. Although atmospheric science has so far been slow to harness the technology,
Outlier detection in sensed data using statistical learning models for IoT
An IoT architecture to detect the occurrence of both Error and Event in a forest environment with the help of four statistical models, i.e., Classification and Regression Trees (CART), Random Forest (RF), Gradient Boosting Machine (GBM) and Linear Discriminant Analysis (LDA).
Bayesian Inference for Large Scale Image Classification
ATMC, an adaptive noise MCMC algorithm that estimates and is able to sample from the posterior of a neural network, is introduced and is shown to be intrinsically robust to overfitting on the training data and to provide a better calibrated measure of uncertainty compared to the optimization baseline.
Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
This work proposes an alternative to Bayesian NNs that is simple to implement, readily parallelizable, requires very little hyperparameter tuning, and yields high quality predictive uncertainty estimates.
Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
A new theoretical framework is developed casting dropout training in deep neural networks (NNs) as approximate Bayesian inference in deep Gaussian processes, which mitigates the problem of representing uncertainty in deep learning without sacrificing either computational complexity or test accuracy.
Outlier detection approaches for wireless sensor networks: A survey