English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Only increasing accuracy without considering uncertainty may negativelyimpact Deep Neural Network (DNN) decision-making and decrease its reliability.This paper proposes five combined preprocessing and post-processing methods fortime-series binary classification problems that simultaneously increase theaccuracy and reliability of DNN outputs applied in a 5G UAV security dataset.These techniques use DNN outputs as input parameters and process them indifferent ways. Two methods use a well-known Machine Learning (ML) algorithm asa complement, and the other three use only confidence values that the DNNestimates. We compare seven different metrics, such as the Expected CalibrationError (ECE), Maximum Calibration Error (MCE), Mean Confidence (MC), MeanAccuracy (MA), Normalized Negative Log Likelihood (NLL), Brier Score Loss(BSL), and Reliability Score (RS) and the tradeoffs between them to evaluatethe proposed hybrid algorithms. First, we show that the eXtreme GradientBoosting (XGB) classifier might not be reliable for binary classification underthe conditions this work presents. Second, we demonstrate that at least one ofthe potential methods can achieve better results than the classification in theDNN softmax layer. Finally, we show that the prospective methods may improveaccuracy and reliability with better uncertainty calibration based on theassumption that the RS determines the difference between MC and MA metrics, andthis difference should be zero to increase reliability. For example, Method 3presents the best RS of 0.65 even when compared to the XGB classifier, whichachieves RS of 7.22.

Accurate and Reliable Methods for 5G UAV Jamming Identification With Calibrated Uncertainty