Using Machine Learning Techniques to Recover Prismatic Cirrus Ice Crystal Size from 2-Dimensional Light Scattering Patterns

Daniel Priori,; Giseli de Sousa; Mauro Roisenberg; Chris Stopford; Evelyn Hesse; Neil Davey; Yi Sun

doi:10.1007/978-3-319-44781-0_44

Using Machine Learning Techniques to Recover Prismatic Cirrus Ice Crystal Size from 2-Dimensional Light Scattering Patterns

Daniel Priori,, Giseli de Sousa, Mauro Roisenberg, Chris Stopford, Evelyn Hesse, Neil Davey, Yi Sun

Research output: Contribution to journal › Conference article › peer-review

1 Citation (Scopus)

Abstract

In this paper, we present a prediction model developed to
identify particles size of ice crystals in clouds. The proposed model combines a Feed Forward Multi-Layer Perceptron neural network withBayesian regularization backpropagation and other machine learning techniques for feature reduction with Principal Component Analysis androtation invariance with Fast Fourier Transform. The proposed solution is capable of predicting the particle sizes with normalized mean squared error around 0.007. However, the proposed network model is not able topredict the size of very small particles (between 3 and 10 µm size) with the same precision as for the larger particles. Therefore, in this work we also discuss some possible reasons for this problem and suggest future points that need to be analysed.

Original language	English
Pages (from-to)	372-379
Number of pages	8
Journal	Lecture Notes in Computer Science (LNCS)
Volume	9887
DOIs	https://doi.org/10.1007/978-3-319-44781-0_44
Publication status	E-pub ahead of print - 13 Aug 2016
Event	25th International Conference on Artificial Neural Networks - Barcelona Tech, Universitat Politecnica de Catalunya, Barcelona, Spain Duration: 6 Sept 2016 → 9 Sept 2016 http://icann2016.org/

Keywords

2d light scattering pattern
Atmospheric particle
size prediction
Fast Fourier Transform
Neural network regression

Access to Document

10.1007/978-3-319-44781-0_44

Cite this

@article{4ee36534ae3e4a02a02798c6ac4225e2,

title = "Using Machine Learning Techniques to Recover Prismatic Cirrus Ice Crystal Size from 2-Dimensional Light Scattering Patterns",

abstract = "In this paper, we present a prediction model developed toidentify particles size of ice crystals in clouds. The proposed model combines a Feed Forward Multi-Layer Perceptron neural network withBayesian regularization backpropagation and other machine learning techniques for feature reduction with Principal Component Analysis androtation invariance with Fast Fourier Transform. The proposed solution is capable of predicting the particle sizes with normalized mean squared error around 0.007. However, the proposed network model is not able topredict the size of very small particles (between 3 and 10 µm size) with the same precision as for the larger particles. Therefore, in this work we also discuss some possible reasons for this problem and suggest future points that need to be analysed.",

keywords = "2d light scattering pattern, Atmospheric particle, size prediction, Fast Fourier Transform, Neural network regression",

author = "Daniel Priori, and {de Sousa}, Giseli and Mauro Roisenberg and Chris Stopford and Evelyn Hesse and Neil Davey and Yi Sun",

note = "Daniel Priori, Giseli de Sousa, Mauro Roisenberg, Chris Stopford, Evelyn Hesse, Neil Davey and Yi Sun, 'Using Machine Learning Techniques to Recover Prismatic Cirrus Ice Crystal Size from 2-Dimensional Light Scattering Patterns', in Alessandro E. P. Villa, Paolo Masulli, and Antonio J. Pons Rivero eds., Proceedings of Artificial Neural Networks and Machine Learning – ICANN 2016: 25th International Conference on Artificial Neural Networks. Universitat Politecnica de Catalunya, Barcelona, Spain, 6- 9 September 2016. ISBN 978-3-319-44780-3, e-ISBN 978-3-319-44781-0; 25th International Conference on Artificial Neural Networks, ICANN ; Conference date: 06-09-2016 Through 09-09-2016",

year = "2016",

month = aug,

day = "13",

doi = "10.1007/978-3-319-44781-0_44",

language = "English",

volume = "9887 ",

pages = "372--379",

journal = "Lecture Notes in Computer Science (LNCS)",

issn = "0302-9743",

publisher = "Springer Nature Link",

url = "http://icann2016.org/",

}

TY - JOUR

T1 - Using Machine Learning Techniques to Recover Prismatic Cirrus Ice Crystal Size from 2-Dimensional Light Scattering Patterns

AU - Priori,, Daniel

AU - de Sousa, Giseli

AU - Roisenberg, Mauro

AU - Stopford, Chris

AU - Hesse, Evelyn

AU - Davey, Neil

AU - Sun, Yi

N1 - Daniel Priori, Giseli de Sousa, Mauro Roisenberg, Chris Stopford, Evelyn Hesse, Neil Davey and Yi Sun, 'Using Machine Learning Techniques to Recover Prismatic Cirrus Ice Crystal Size from 2-Dimensional Light Scattering Patterns', in Alessandro E. P. Villa, Paolo Masulli, and Antonio J. Pons Rivero eds., Proceedings of Artificial Neural Networks and Machine Learning – ICANN 2016: 25th International Conference on Artificial Neural Networks. Universitat Politecnica de Catalunya, Barcelona, Spain, 6- 9 September 2016. ISBN 978-3-319-44780-3, e-ISBN 978-3-319-44781-0

PY - 2016/8/13

Y1 - 2016/8/13

N2 - In this paper, we present a prediction model developed toidentify particles size of ice crystals in clouds. The proposed model combines a Feed Forward Multi-Layer Perceptron neural network withBayesian regularization backpropagation and other machine learning techniques for feature reduction with Principal Component Analysis androtation invariance with Fast Fourier Transform. The proposed solution is capable of predicting the particle sizes with normalized mean squared error around 0.007. However, the proposed network model is not able topredict the size of very small particles (between 3 and 10 µm size) with the same precision as for the larger particles. Therefore, in this work we also discuss some possible reasons for this problem and suggest future points that need to be analysed.

AB - In this paper, we present a prediction model developed toidentify particles size of ice crystals in clouds. The proposed model combines a Feed Forward Multi-Layer Perceptron neural network withBayesian regularization backpropagation and other machine learning techniques for feature reduction with Principal Component Analysis androtation invariance with Fast Fourier Transform. The proposed solution is capable of predicting the particle sizes with normalized mean squared error around 0.007. However, the proposed network model is not able topredict the size of very small particles (between 3 and 10 µm size) with the same precision as for the larger particles. Therefore, in this work we also discuss some possible reasons for this problem and suggest future points that need to be analysed.

KW - 2d light scattering pattern

KW - Atmospheric particle

KW - size prediction

KW - Fast Fourier Transform

KW - Neural network regression

U2 - 10.1007/978-3-319-44781-0_44

DO - 10.1007/978-3-319-44781-0_44

M3 - Conference article

SN - 0302-9743

VL - 9887

SP - 372

EP - 379

JO - Lecture Notes in Computer Science (LNCS)

JF - Lecture Notes in Computer Science (LNCS)

T2 - 25th International Conference on Artificial Neural Networks

Y2 - 6 September 2016 through 9 September 2016

ER -

Using Machine Learning Techniques to Recover Prismatic Cirrus Ice Crystal Size from 2-Dimensional Light Scattering Patterns

Abstract

Keywords

Access to Document

Fingerprint

Cite this