Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques
"Gravitatonal waves, the seed of the 2015 Nobel’s prize are the cause of several complex celestial phenomena that is non-observable for the naked eye. Their identification, classification and study is s(ll a handmade work which is s(ll nascent. There has been several approaches to produce novel...
| Autor principal: | |
|---|---|
| Otros Autores: | |
| Formato: | Trabajo final de especialización |
| Lenguaje: | Inglés |
| Publicado: |
2021
|
| Materias: | |
| Acceso en línea: | http://ri.itba.edu.ar/handle/123456789/3351 |
| Aporte de: |
| id |
I32-R138-123456789-3351 |
|---|---|
| record_format |
dspace |
| spelling |
I32-R138-123456789-33512022-12-07T15:24:06Z Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques Comparativa de rendimiento de clasificacion de espectrogramas de ondas gravitacionales mediante la utilizacion de tecnicas de machine learning Martínez, Ezequiel H. Ramele, Rodrigo APRENDIZAJE AUTOMATICO PROCESAMIENTO DE SEÑALES CLASIFICACION ONDAS GRAVITACIONALES "Gravitatonal waves, the seed of the 2015 Nobel’s prize are the cause of several complex celestial phenomena that is non-observable for the naked eye. Their identification, classification and study is s(ll a handmade work which is s(ll nascent. There has been several approaches to produce novel tools to aid the scientists behind the discovery of these deep space events. One of the most thrilling examples has been the usage of artificial intelligence classification to aid in the preiden identification of certain signals. We took one of these tools, Gravity Spy, and study its base paper, trying to reproduce some of their classification results using the very same base dataset. This research aims to compare the results obtained from the original paper, with a binary classification approach and several different algorithms taken from the knowledge base of machine learning and deep learning, alike. We confirmed the original paper results and obtained a new approach for the same solution. In this study we trained several models that could be used for further development of an eventual alternative engines for gravita(onal waves signal’s classification or any other sort of signal heavily influenced by noise and analysed by spectrograms." Trabajo Final Ciencia de Datos (especialización) - Instituto Tecnológico de Buenos Aires, Buenos Aires, 2020 2021-01-26T14:51:23Z 2021-01-26T14:51:23Z 2020-09-06 Trabajo final de especialización http://ri.itba.edu.ar/handle/123456789/3351 en application/pdf |
| institution |
Instituto Tecnológico de Buenos Aires (ITBA) |
| institution_str |
I-32 |
| repository_str |
R-138 |
| collection |
Repositorio Institucional Instituto Tecnológico de Buenos Aires (ITBA) |
| language |
Inglés |
| topic |
APRENDIZAJE AUTOMATICO PROCESAMIENTO DE SEÑALES CLASIFICACION ONDAS GRAVITACIONALES |
| spellingShingle |
APRENDIZAJE AUTOMATICO PROCESAMIENTO DE SEÑALES CLASIFICACION ONDAS GRAVITACIONALES Martínez, Ezequiel H. Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques |
| topic_facet |
APRENDIZAJE AUTOMATICO PROCESAMIENTO DE SEÑALES CLASIFICACION ONDAS GRAVITACIONALES |
| description |
"Gravitatonal waves, the seed of the 2015 Nobel’s prize are the cause of several complex celestial phenomena that is non-observable for the naked eye. Their identification, classification and study is s(ll a handmade work which is s(ll nascent. There has been several approaches to produce novel tools to aid the scientists behind the discovery of these deep space events. One of the most thrilling examples has been the usage of artificial intelligence classification to aid in the preiden identification of certain signals. We took one of these tools, Gravity Spy, and study its base paper, trying to reproduce some of their classification results using the very same base dataset.
This research aims to compare the results obtained from the original paper, with a binary classification approach and several different algorithms taken from the knowledge base of machine learning and deep learning, alike. We confirmed the original paper results and obtained a new approach for the same solution. In this study we trained several models that could be used for
further development of an eventual alternative engines for gravita(onal waves signal’s classification or any other sort of signal heavily influenced by noise and analysed by spectrograms." |
| author2 |
Ramele, Rodrigo |
| author_facet |
Ramele, Rodrigo Martínez, Ezequiel H. |
| format |
Trabajo final de especialización |
| author |
Martínez, Ezequiel H. |
| author_sort |
Martínez, Ezequiel H. |
| title |
Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques |
| title_short |
Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques |
| title_full |
Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques |
| title_fullStr |
Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques |
| title_full_unstemmed |
Analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques |
| title_sort |
analysis and benchmarking for gravitational waves spectrogram’s classification by usage of machine learning techniques |
| publishDate |
2021 |
| url |
http://ri.itba.edu.ar/handle/123456789/3351 |
| work_keys_str_mv |
AT martinezezequielh analysisandbenchmarkingforgravitationalwavesspectrogramsclassificationbyusageofmachinelearningtechniques AT martinezezequielh comparativaderendimientodeclasificaciondeespectrogramasdeondasgravitacionalesmediantelautilizaciondetecnicasdemachinelearning |
| _version_ |
1766093671112900608 |