Sorry, you need to enable JavaScript to visit this website.

Pattern recognition and classification (MLR-PATT)

Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry


The Viterbi algorithm and its pruning variant, are some of the most frequently used algorithms in communications and speech recognition. There has been extended research on improving the algorithms’ computational complexity, however work trying to interpret their nonlinear structure and geometry has been limited. In this work we analyse the Viterbi algorithm in the field of tropical (min-plus) algebra, and we utilize its pruning variant in order to define a polytope. Then, we interpret certain faces of the polytope as the most probable states of the algorithm.

Paper Details

Authors:
Petros Maragos
Submitted On:
22 June 2018 - 8:25am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Poster

(69 downloads)

Subscribe

[1] Petros Maragos, "Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3292. Accessed: Sep. 18, 2018.
@article{3292-18,
url = {http://sigport.org/3292},
author = {Petros Maragos },
publisher = {IEEE SigPort},
title = {Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry},
year = {2018} }
TY - EJOUR
T1 - Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry
AU - Petros Maragos
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3292
ER -
Petros Maragos. (2018). Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry. IEEE SigPort. http://sigport.org/3292
Petros Maragos, 2018. Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry. Available at: http://sigport.org/3292.
Petros Maragos. (2018). "Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry." Web.
1. Petros Maragos. Analysis of the Viterbi Algorithm Using Tropical Algebra and Geometry [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3292

PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS


We consider the problem of predicting power outages in an electrical power grid due to hazards produced by convective storms. These storms produce extreme weather phenomena such as intense wind, tornadoes and lightning over a small area. In this paper, we discuss the application of state-of-the-art machine learning techniques, such as random forest classifiers and deep neural networks, to predict the amount of damage caused by storms.

Paper Details

Authors:
Joonas Karjalainen
Submitted On:
29 May 2018 - 3:22am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

SASSE_poster_IEEE_DSW_2018.pdf

(47 downloads)

Subscribe

[1] Joonas Karjalainen, "PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3211. Accessed: Sep. 18, 2018.
@article{3211-18,
url = {http://sigport.org/3211},
author = {Joonas Karjalainen },
publisher = {IEEE SigPort},
title = {PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS},
year = {2018} }
TY - EJOUR
T1 - PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS
AU - Joonas Karjalainen
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3211
ER -
Joonas Karjalainen. (2018). PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS. IEEE SigPort. http://sigport.org/3211
Joonas Karjalainen, 2018. PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS. Available at: http://sigport.org/3211.
Joonas Karjalainen. (2018). "PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS." Web.
1. Joonas Karjalainen. PREDICTING ELECTRICITY OUTAGES CAUSED BY CONVECTIVE STORMS [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3211

CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING

Paper Details

Authors:
Yongjian Xue, Pierre Beauseroy
Submitted On:
24 April 2018 - 4:44am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Poster_Yongjian_ICASSP.pdf

(61 downloads)

Subscribe

[1] Yongjian Xue, Pierre Beauseroy, "CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3164. Accessed: Sep. 18, 2018.
@article{3164-18,
url = {http://sigport.org/3164},
author = {Yongjian Xue; Pierre Beauseroy },
publisher = {IEEE SigPort},
title = {CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING},
year = {2018} }
TY - EJOUR
T1 - CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING
AU - Yongjian Xue; Pierre Beauseroy
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3164
ER -
Yongjian Xue, Pierre Beauseroy. (2018). CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING. IEEE SigPort. http://sigport.org/3164
Yongjian Xue, Pierre Beauseroy, 2018. CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING. Available at: http://sigport.org/3164.
Yongjian Xue, Pierre Beauseroy. (2018). "CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING." Web.
1. Yongjian Xue, Pierre Beauseroy. CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3164

CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING

Paper Details

Authors:
Yongjian Xue, Pierre Beauseroy
Submitted On:
24 April 2018 - 4:44am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Poster_Yongjian_ICASSP.pdf

(55 downloads)

Subscribe

[1] Yongjian Xue, Pierre Beauseroy, "CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/3163. Accessed: Sep. 18, 2018.
@article{3163-18,
url = {http://sigport.org/3163},
author = {Yongjian Xue; Pierre Beauseroy },
publisher = {IEEE SigPort},
title = {CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING},
year = {2018} }
TY - EJOUR
T1 - CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING
AU - Yongjian Xue; Pierre Beauseroy
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/3163
ER -
Yongjian Xue, Pierre Beauseroy. (2018). CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING. IEEE SigPort. http://sigport.org/3163
Yongjian Xue, Pierre Beauseroy, 2018. CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING. Available at: http://sigport.org/3163.
Yongjian Xue, Pierre Beauseroy. (2018). "CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING." Web.
1. Yongjian Xue, Pierre Beauseroy. CONSTANT FALSE ALARM RATE FOR ONLINE ONE CLASS SVM LEARNING [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/3163

Clustering of data with missing entries


The analysis of large datasets is often complicated by the presence of missing entries, mainly because most of the current machine learning algorithms are designed to work with full data. The main focus of this work is to introduce a clustering

Paper Details

Authors:
Sunrita Poddar, Mathews Jacob
Submitted On:
14 April 2018 - 8:11pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

clusteringMissingEntries

(53 downloads)

icassp18poster.pdf

(74 downloads)

Subscribe

[1] Sunrita Poddar, Mathews Jacob, "Clustering of data with missing entries", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/2858. Accessed: Sep. 18, 2018.
@article{2858-18,
url = {http://sigport.org/2858},
author = {Sunrita Poddar; Mathews Jacob },
publisher = {IEEE SigPort},
title = {Clustering of data with missing entries},
year = {2018} }
TY - EJOUR
T1 - Clustering of data with missing entries
AU - Sunrita Poddar; Mathews Jacob
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/2858
ER -
Sunrita Poddar, Mathews Jacob. (2018). Clustering of data with missing entries. IEEE SigPort. http://sigport.org/2858
Sunrita Poddar, Mathews Jacob, 2018. Clustering of data with missing entries. Available at: http://sigport.org/2858.
Sunrita Poddar, Mathews Jacob. (2018). "Clustering of data with missing entries." Web.
1. Sunrita Poddar, Mathews Jacob. Clustering of data with missing entries [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/2858

TV-SVM: Support Vector Machine with Total Variational Regularization


To leverage the spatial relationship of lattice data, such as images, we introduce total variational (TV) regularization into support vector machines (SVM), called TV-SVM. TV-SVM encourages local smoothness and sparsity in gradient domain of the learned parameters. TV-SVM is optimized via the alternating direction method of multipliers (ADMM) algorithm and is significantly better than (Linear) SVM for image classifications.

Paper Details

Authors:
Zhendong Zhang,Cheolkon Jung
Submitted On:
13 April 2018 - 12:30pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP2018poster_TVSVM_final.pdf

(97 downloads)

Subscribe

[1] Zhendong Zhang,Cheolkon Jung, "TV-SVM: Support Vector Machine with Total Variational Regularization", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/2728. Accessed: Sep. 18, 2018.
@article{2728-18,
url = {http://sigport.org/2728},
author = {Zhendong Zhang;Cheolkon Jung },
publisher = {IEEE SigPort},
title = {TV-SVM: Support Vector Machine with Total Variational Regularization},
year = {2018} }
TY - EJOUR
T1 - TV-SVM: Support Vector Machine with Total Variational Regularization
AU - Zhendong Zhang;Cheolkon Jung
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/2728
ER -
Zhendong Zhang,Cheolkon Jung. (2018). TV-SVM: Support Vector Machine with Total Variational Regularization. IEEE SigPort. http://sigport.org/2728
Zhendong Zhang,Cheolkon Jung, 2018. TV-SVM: Support Vector Machine with Total Variational Regularization. Available at: http://sigport.org/2728.
Zhendong Zhang,Cheolkon Jung. (2018). "TV-SVM: Support Vector Machine with Total Variational Regularization." Web.
1. Zhendong Zhang,Cheolkon Jung. TV-SVM: Support Vector Machine with Total Variational Regularization [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/2728

TV-SVM: Support Vector Machine with Total Variational Regularization


To leverage the spatial relationship of lattice data, such as images, we introduce total variational (TV) regularization into support vector machines (SVM), called TV-SVM. TV-SVM encourages local smoothness and sparsity in gradient domain of the learned parameters. TV-SVM is optimized via the alternating direction method of multipliers (ADMM) algorithm and is significantly better than (Linear) SVM for image classifications.

Paper Details

Authors:
Zhendong Zhang,Cheolkon Jung
Submitted On:
13 April 2018 - 12:22pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP2018poster_TVSVM_final.pdf

(76 downloads)

Subscribe

[1] Zhendong Zhang,Cheolkon Jung, "TV-SVM: Support Vector Machine with Total Variational Regularization", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/2724. Accessed: Sep. 18, 2018.
@article{2724-18,
url = {http://sigport.org/2724},
author = {Zhendong Zhang;Cheolkon Jung },
publisher = {IEEE SigPort},
title = {TV-SVM: Support Vector Machine with Total Variational Regularization},
year = {2018} }
TY - EJOUR
T1 - TV-SVM: Support Vector Machine with Total Variational Regularization
AU - Zhendong Zhang;Cheolkon Jung
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/2724
ER -
Zhendong Zhang,Cheolkon Jung. (2018). TV-SVM: Support Vector Machine with Total Variational Regularization. IEEE SigPort. http://sigport.org/2724
Zhendong Zhang,Cheolkon Jung, 2018. TV-SVM: Support Vector Machine with Total Variational Regularization. Available at: http://sigport.org/2724.
Zhendong Zhang,Cheolkon Jung. (2018). "TV-SVM: Support Vector Machine with Total Variational Regularization." Web.
1. Zhendong Zhang,Cheolkon Jung. TV-SVM: Support Vector Machine with Total Variational Regularization [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/2724

TV-SVM: Support Vector Machine with Total Variational Regularization


To leverage the spatial relationship of lattice data, such as images, we introduce total variational (TV) regularization into support vector machines (SVM), called TV-SVM. TV-SVM encourages local smoothness and sparsity in gradient domain of the learned parameters. TV-SVM is optimized via the alternating direction method of multipliers (ADMM) algorithm and is significantly better than (Linear) SVM for image classifications.

Paper Details

Authors:
Zhendong Zhang,Cheolkon Jung
Submitted On:
13 April 2018 - 12:22pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP2018poster_TVSVM_final.pdf

(80 downloads)

Subscribe

[1] Zhendong Zhang,Cheolkon Jung, "TV-SVM: Support Vector Machine with Total Variational Regularization", IEEE SigPort, 2018. [Online]. Available: http://sigport.org/2623. Accessed: Sep. 18, 2018.
@article{2623-18,
url = {http://sigport.org/2623},
author = {Zhendong Zhang;Cheolkon Jung },
publisher = {IEEE SigPort},
title = {TV-SVM: Support Vector Machine with Total Variational Regularization},
year = {2018} }
TY - EJOUR
T1 - TV-SVM: Support Vector Machine with Total Variational Regularization
AU - Zhendong Zhang;Cheolkon Jung
PY - 2018
PB - IEEE SigPort
UR - http://sigport.org/2623
ER -
Zhendong Zhang,Cheolkon Jung. (2018). TV-SVM: Support Vector Machine with Total Variational Regularization. IEEE SigPort. http://sigport.org/2623
Zhendong Zhang,Cheolkon Jung, 2018. TV-SVM: Support Vector Machine with Total Variational Regularization. Available at: http://sigport.org/2623.
Zhendong Zhang,Cheolkon Jung. (2018). "TV-SVM: Support Vector Machine with Total Variational Regularization." Web.
1. Zhendong Zhang,Cheolkon Jung. TV-SVM: Support Vector Machine with Total Variational Regularization [Internet]. IEEE SigPort; 2018. Available from : http://sigport.org/2623

Scattering Features for Multimodal Gait Recognition


Gait.pdf

PDF icon Gait.pdf (150 downloads)

Paper Details

Authors:
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton
Submitted On:
25 November 2017 - 8:19pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Gait.pdf

(150 downloads)

Subscribe

[1] Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton, "Scattering Features for Multimodal Gait Recognition", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/2369. Accessed: Sep. 18, 2018.
@article{2369-17,
url = {http://sigport.org/2369},
author = {Srdan Kitic;Gilles Puy;Patrick Perez;Philippe Gilberton },
publisher = {IEEE SigPort},
title = {Scattering Features for Multimodal Gait Recognition},
year = {2017} }
TY - EJOUR
T1 - Scattering Features for Multimodal Gait Recognition
AU - Srdan Kitic;Gilles Puy;Patrick Perez;Philippe Gilberton
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/2369
ER -
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton. (2017). Scattering Features for Multimodal Gait Recognition. IEEE SigPort. http://sigport.org/2369
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton, 2017. Scattering Features for Multimodal Gait Recognition. Available at: http://sigport.org/2369.
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton. (2017). "Scattering Features for Multimodal Gait Recognition." Web.
1. Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton. Scattering Features for Multimodal Gait Recognition [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/2369

Scattering Features for Multimodal Gait Recognition


Gait.pdf

PDF icon Gait.pdf (138 downloads)

Paper Details

Authors:
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton
Submitted On:
25 November 2017 - 8:19pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Gait.pdf

(138 downloads)

Subscribe

[1] Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton, "Scattering Features for Multimodal Gait Recognition", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/2368. Accessed: Sep. 18, 2018.
@article{2368-17,
url = {http://sigport.org/2368},
author = {Srdan Kitic;Gilles Puy;Patrick Perez;Philippe Gilberton },
publisher = {IEEE SigPort},
title = {Scattering Features for Multimodal Gait Recognition},
year = {2017} }
TY - EJOUR
T1 - Scattering Features for Multimodal Gait Recognition
AU - Srdan Kitic;Gilles Puy;Patrick Perez;Philippe Gilberton
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/2368
ER -
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton. (2017). Scattering Features for Multimodal Gait Recognition. IEEE SigPort. http://sigport.org/2368
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton, 2017. Scattering Features for Multimodal Gait Recognition. Available at: http://sigport.org/2368.
Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton. (2017). "Scattering Features for Multimodal Gait Recognition." Web.
1. Srdan Kitic,Gilles Puy,Patrick Perez,Philippe Gilberton. Scattering Features for Multimodal Gait Recognition [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/2368

Pages