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

In recent years, machine learning (ML) and neural networks (NNs) have gainedwidespread use and attention across various domains, particularly intransportation for achieving autonomy, including the emergence of flying taxisfor urban air mobility (UAM). However, concerns about certification have comeup, compelling the development of standardized processes encompassing theentire ML and NN pipeline. This paper delves into the inference stage and therequisite hardware, highlighting the challenges associated with IEEE 754floating-point arithmetic and proposing alternative number representations. Byevaluating diverse summation and dot product algorithms, we aim to mitigateissues related to non-associativity. Additionally, our exploration offixed-point arithmetic reveals its advantages over floating-point methods,demonstrating significant hardware efficiencies. Employing an empiricalapproach, we ascertain the optimal bit-width necessary to attain an acceptablelevel of accuracy, considering the inherent complexity of bit-widthoptimization.

Efficient and Mathematically Robust Operations for Certified Neural Networks Inference