<b>NeuralNetTools</b>: Visualization and Analysis Tools for Neural Networks

Beck, Marcus W.

doi:10.18637/jss.v085.i11

Cited by 228 publications

(160 citation statements)

References 32 publications

Supporting

Mentioning

157

Contrasting

Unclassified

Order By: Relevance

“…for two hidden layers, testing 40 total nodes would have tested 10, 20 and 30 nodes in layer 1 and 30, 20 and 10 nodes in layer 2, respectively. The relative variable importance was computed using the olden function in the R package NeuralNetTools [24].…”

Section: (E) Population Modelling and Deep Learningmentioning

confidence: 99%

Ancient mitogenomes show plateau populations from last 5200 years partially contributed to present-day Tibetans

Ding

Wang

et al. 2020

Proc. R. Soc. B.

View full text Add to dashboard Cite

The clarification of the genetic origins of present-day Tibetans requires an understanding of their past relationships with the ancient populations of the Tibetan Plateau. Here we successfully sequenced 67 complete mitochondrial DNA genomes of 5200 to 300-year-old humans from the plateau. Apart from identifying two ancient plateau lineages (haplogroups D4j1b and M9a1a1c1b1a) that suggest some ancestors of Tibetans came from low-altitude areas 4750 to 2775 years ago and that some were involved in an expansion of people moving between high-altitude areas 2125 to 1100 years ago, we found limited evidence of recent matrilineal continuity on the plateau. Furthermore, deep learning of the ancient data incorporated into simulation models with an accuracy of 97% supports that present-day Tibetan matrilineal ancestry received partial contribution rather than complete continuity from the plateau populations of the last 5200 years.

show abstract

Section: (E) Population Modelling and Deep Learningmentioning

confidence: 99%

Ancient mitogenomes show plateau populations from last 5200 years partially contributed to present-day Tibetans

Ding

Wang

et al. 2020

Proc. R. Soc. B.

View full text Add to dashboard Cite

show abstract

“…The classification accuracy has been measured separately for normal, noisy and a combined data set. Analysis was conducted in R [22] with signal [23] and neuralnet [24] libraries on Intel i7, 8 MB RAM workstation.…”

Section: Resultsmentioning

confidence: 99%

Passive Localization Through Light Flicker Fingerprinting

Munir

Dyo

2019

IEEE Sensors J.

View full text Add to dashboard Cite

In this paper, we show that the flicker waveforms of various CFL and LED lamp models exhibit distinctive waveform patterns due to harmonic distortions of rectifiers and voltage regulators, the key components of modern lamp drivers. We then propose a passive localization technique based on fingerprinting these distortions that occur naturally in indoor environments and thus requires no infrastructure or additional equipment. The novel technique uses principal component analysis (PCA) to extract the most important signal features from the flicker frequency spectra followed by kNN clustering and neural network classifiers to identify a light source based on its flicker signature. The evaluation on 39 flicker patterns collected from 8 residential locations demonstrates that the technique can identify a location within a house with up to 90% accuracy and identify an individual house from a set of houses with an average accuracy of 86.3%.

show abstract

“…; Venables & Ripley ) used in this study is a relatively simple implementation using a feedforward network with a single hidden layer, and includes two main parameters that need to be adjusted to optimise performance: Size , which specifies the number of nodes in the hidden layer, and Decay , which is a regularisation parameter applied to weighting factors to minimise overfitting. Examples of NN structures applied to the obsidian data are shown in Supporting Information Figure (see also Beck ).…”

Section: Methodsmentioning

confidence: 99%

On provenance studies of New Zealand obsidians: A pXRF‐based geochemical reference dataset and a review of analytical methods

McAlister

2019

Archaeology in Oceania

View full text Add to dashboard Cite

Since the introduction of portable X-ray fluorescence (pXRF) technology to obsidian provenance research, the volume of published studies has increased exponentially. This proliferation of data creates unprecedented analytic opportunities for the development of higher level archaeological theory but, at the same time, presents methodological challenges. Here, it is argued that rigorous analytic procedures are required to keep pXRF-based provenance studies in line with the best scientific practice; specifically, the generation of accurate compositional data, the compilation of comprehensive reference databases for robust source characterisation and the development of appropriate methods for artefact assignment are essential components of a reliable protocol. To these ends, a reference dataset for New Zealand obsidians containing over 1500 samples is reported in this paper, previous source assignment methods are reviewed and a selection of multivariate approaches are evaluated. RÉSUMÉLesétudes de provenance des artefacts en obsidienne se sont développées de manière exponentielle avec l'introduction de la spectrométrie portable XRF (pXRF). Cette profusion de données crée des opportunités analytiques sans précédent qui permettent d'importants développements théoriques, mais présenteégalement des défis méthodologiques. Nous montrons ici qu'il est crucial de mettre en place et d'utiliser des procédures analytiques rigoureuses afin de conserver la qualité desétudes de provenance basées sur le pXRF. La fiabilité des protocoles est notamment liéeà la précision des données de composition, la compilation de bases de données de référence complètes pour une caractérisation robuste des sources, et le développement de méthodes appropriées qui permettent l'attribution de source pour chaque artefact. Dans cet objectif, cet article présente un ensemble de données de référence pour les obsidiennes de Nouvelle-Zélande contenant plus de 1500échantillons, et propose uneévaluation critique des méthodes d'attribution des sources ainsi que de plusieurs approches statistiques multivariées couramment utilisées dans lesétudes de provenance.

show abstract

NeuralNetTools: Visualization and Analysis Tools for Neural Networks

Cited by 228 publications

References 32 publications

Ancient mitogenomes show plateau populations from last 5200 years partially contributed to present-day Tibetans

Ancient mitogenomes show plateau populations from last 5200 years partially contributed to present-day Tibetans

Passive Localization Through Light Flicker Fingerprinting

On provenance studies of New Zealand obsidians: A pXRF‐based geochemical reference dataset and a review of analytical methods

Contact Info

Product

Resources

About