Dados de pesquisa de área ampla podem ser usados ​​para estudar a estrutura em grande escala do Universo por meio de media§aµes de padraµes de lentes gravitacionais.
![](data:image/jpeg;base64,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)
Visualização artastica desta pesquisa. Usando a análise de dados orientada por IA para remover o ruado e encontrar a forma real do universo. Crédito: The Institute of Statistical Mathematics
Astra´nomos japoneses desenvolveram uma nova técnica de inteligaªncia artificial (IA) para remover ruado em dados astrona´micos devido a variações aleata³rias nas formas das gala¡xias. Apa³s extenso treinamento e testes em grandes dados fictacios criados por simulações de supercomputadores, eles aplicaram esta nova ferramenta aos dados reais do Telescópio Subaru do Japa£o e descobriram que a distribuição de massa derivada do uso deste manãtodo éconsistente com os modelos atualmente aceitos do Universo. Esta éuma nova ferramenta poderosa para analisar grandes volumes de dados de pesquisas astrona´micas atuais e planejadas.
Dados de pesquisa de área ampla podem ser usados ​​para estudar a estrutura em grande escala do Universo por meio de medições de padraµes de lentes gravitacionais. Em lentes gravitacionais, a gravidade de um objeto em primeiro plano, como um aglomerado de gala¡xias , pode distorcer a imagem de um objeto em segundo plano, como uma gala¡xia mais distante. Alguns exemplos de lentes gravitacionais são a³bvios, como o "Olho de Ha³rus". A estrutura em grande escala, consistindo principalmente de misteriosa matéria "escura", pode distorcer as formas de gala¡xias distantes também, mas o efeito de lente esperado ésutil. A média de muitas gala¡xias em uma área énecessa¡ria para criar um mapa das distribuições de matéria escura em primeiro plano.
"Esta pesquisa mostra os benefacios de combinar diferentes tipos de pesquisa: observações, simulações e análise de dados de IA ", diz Masato Shirasaki, o lider da equipe, "Nesta era de big data , precisamos ultrapassar as fronteiras tradicionais entre especialidades e usar todas as ferramentas disponíveis para entender os dados. Se pudermos fazer isso, abriremos novos campos na astronomia e outras ciências. "
Mas essa técnica de olhar para muitas imagens de gala¡xias éum problema; algumas gala¡xias são inatamente um pouco engraçadas. a‰ difacil distinguir entre uma imagem de gala¡xia distorcida por lentes gravitacionais e uma gala¡xia que estãorealmente distorcida. Isso éconhecido como ruado de forma e éum dos fatores limitantes na pesquisa que estuda a estrutura em grande escala do Universo.
Esquema da inteligaªncia artificial utilizada neste estudo, uma rede gerativa adversarial
(GAN). A primeira rede, chamada de gerador de imagem G, estima e produz um mapa de
lente sem ruado a partir de um mapa de lente com ruado. A segunda rede, o discriminador
de imagem D, compara o mapa de lentes criado por G com o verdadeiro mapa de lentes
sem ruado e identifica a imagem criada por G como falsa. Ao inserir um grande número
de pares de mapas de lente barulhentos / sem ruado nas duas redes, G étreinado para
fazer mapas de lentes mais pra³ximos dos originais e D étreinado para localizar com
mais precisão as falsificações feitas por G. Neste estudo , 25.000 pares de mapas de
lentes sem ruado e barulhentos obtidos de simulações numanãricas usando ATERUI II
foram usados ​​para criar uma rede esta¡vel. Finalmente, um gerador de imagem
treinado G estima um mapa de lentes sem ruado baseado no mapa de lentes
observacionais ruidosas realmente observado. Crédito: NAOJ
Para compensar o ruado da forma, uma equipe de astrônomos japoneses usou pela primeira vez o ATERUI II, o supercomputador mais poderoso do mundo dedicado a astronomia, para gerar 25.000 cata¡logos de gala¡xias simulados com base em dados reais do Telescópio Subaru. Eles então adicionaram ruado realista a esses conjuntos de dados artificiais perfeitamente conhecidos e treinaram uma IA para recuperar estatisticamente a matéria escura da lente dos dados simulados.
Apa³s o treinamento, a IA foi capaz de recuperar detalhes finos anteriormente inobserva¡veis, ajudando a melhorar nossa compreensão da matéria escura ca³smica. Então, usando esta IA em dados reais cobrindo 21 graus quadrados do canãu, a equipe encontrou uma distribuição da massa do primeiro plano consistente com o modelo cosmola³gico padra£o.
"Esta pesquisa mostra os benefacios de combinar diferentes tipos de pesquisa: observações, simulações e análise de dados de IA ", diz Masato Shirasaki, o lider da equipe, "Nesta era de big data , precisamos ultrapassar as fronteiras tradicionais entre especialidades e usar todas as ferramentas disponíveis para entender os dados. Se pudermos fazer isso, abriremos novos campos na astronomia e outras ciências. "