O Andrew Gelman , professor de estatastica na Columbia, e Aki Vehtari , professor de ciência da computaa§a£o na Universidade Aalto da Finla¢ndia, publicaram uma lista das ideias estatasticas mais importantes dos últimos 50 anos.
![](data:image/jpeg;base64,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)
Reprodução
Se vocêjá pediu ajuda ao Siri ou Alexa, ou gerou um autorretrato no estilo de um pintor da Renascena§a, vocêinteragiu com o aprendizado profundo, uma forma de inteligaªncia artificial que extrai padraµes de montanhas de dados para fazer previsaµes. Embora aprendizado profundo e IA tenham se tornado termos familiares, os avanços nas estatasticas que alimentaram essa revolução são menos conhecidos. Em um artigo recente, Andrew Gelman , professor de estatastica na Columbia, e Aki Vehtari , professor de ciência da computação na Universidade Aalto da Finla¢ndia, publicaram uma lista das ideias estatasticas mais importantes dos últimos 50 anos.
Abaixo, Gelman e Vehtari decompõem a lista para aqueles que podem ter cochilado por meio da Estatastica 101. Cada ideia pode ser vista como um substituto para um subcampo inteiro, dizem eles, com algumas ressalvas: a ciência éincremental; ao destacar essas obras, eles não significam diminuir a importa¢ncia de trabalhos semelhantes e relacionados. Eles também optaram por se concentrar em manãtodos em estatastica e aprendizado de ma¡quina, em vez de avanços igualmente importantes em computação estatastica e ciência da computação e engenharia, que forneceram as ferramentas e o poder de computação para análise e visualização de dados para se tornarem ferramentas prática s do dia a dia. Finalmente, eles se concentraram em manãtodos, embora reconhecendo que os desenvolvimentos na teoria e nos manãtodos são frequentemente motivados por aplicações especaficas.Â
Vaª algo importante que estãofaltando? Tweet-o em @columbiascience e Gelman e Vehtari considerara£o adiciona¡-lo a lista.
Os 10 artigos e livros abaixo todos foram publicados nos últimos 50 anos e estãolistados em ordem cronola³gica.
1. Hirotugu Akaike (1973). Teoria da Informação e uma Extensão do Princapio da Ma¡xima Verossimilhana§a . Anais do Segundo Simpa³sio Internacional de Teoria da Informação.
Este éo artigo que introduziu o termo AIC (originalmente chamado de Critanãrio de Informação, mas agora conhecido como Critanãrio de Informação de Akaike), para avaliar o ajuste de um modelo com base em sua precisão preditiva estimada. AIC foi imediatamente reconhecida como uma ferramenta útil, e este artigo foi um dos vários publicados em meados da década de 1970, colocando a inferaªncia estatastica dentro de uma estrutura preditiva. Agora reconhecemos a validação preditiva como um princapio fundamental em estatastica e aprendizado de ma¡quina. Akaike foi um estatastico aplicado que, na década de 1960, tentou medir a rugosidade das pistas de aeroportos, da mesma forma que os primeiros artigos de Benoit Mandelbrot sobre taxonomia e distribuições de Pareto levaram a seu trabalho posterior na matemática dos fractais.
2. John Tukey (1977). Ana¡lise Explorata³ria de Dados .
Este livro foi extremamente influente e éuma leitura divertida que pode ser digerida de uma são vez. Tradicionalmente, a visualização e a exploração de dados eram consideradas aspectos de baixo grau da estatastica prática ; o fascanio estava em ajustar modelos, provar teoremas e desenvolver as propriedades tea³ricas de procedimentos estatasticos sob várias suposições ou restrições matemáticas. Tukey virou essa ideia de cabea§a para baixo. Ele escreveu sobre ferramentas estatasticas não para confirmar o que já sabaamos (ou pensa¡vamos que sabaamos), e não para rejeitar hipa³teses em que nunca, ou nunca deveraamos, acreditar, mas para descobrir novos e inesperados insights a partir dos dados. Seu trabalho motivou avanços em análise de rede, software e perspectivas tea³ricas que integram confirmação, cratica e descoberta.Â
3. Grace Wahba (1978). Priors impra³prios, suavização de spline e o problema de proteção contra erros de modelo na regressão . Journal of the Royal Statistical Society.
A suavização de spline éuma abordagem para ajustar curvas não paramanãtricas. Outro artigo de Wahba desse período échamado de "Uma curva francesa automa¡tica", referindo-se a uma classe de algoritmos que podem ajustar curvas suaves arbitra¡rias por meio de dados sem overfitting para ruado ou outliers. A ideia pode parecer a³bvia agora, mas foi um grande passo a frente em uma era em que os pontos de partida para o ajuste da curva eram polina´mios, exponenciais e outras formas fixas. Além da aplicabilidade direta de splines, este artigo foi importante teoricamente. Serviu como base para trabalhos posteriores em inferaªncia bayesiana não paramanãtrica ao unificar ideias de regularização de modelos de alta dimensão.
4. Bradley Efron (1979). Manãtodos de bootstrap: outra olhada no Jackknife . Annals of Statistics.
Bootstrapping éum manãtodo para realizar inferaªncia estatastica sem suposições. Os dados puxam-se por seus bootstraps, por assim dizer. Mas vocênão pode fazer inferaªncias sem suposições; o que tornou o bootstrap tão útil e influente éque as suposições vieram implicitamente com o procedimento computacional: a ideia audaciosamente simples de reamostrar os dados. Cada vez que vocêrepete o procedimento estatastico executado nos dados originais. Assim como muitos manãtodos estatasticos dos últimos 50 anos, este se tornou amplamente útil por causa de uma explosão no poder da computação que permitiu que as simulações substituassem a análise matemática.
5. Alan Gelfand e Adrian Smith (1990). Abordagens baseadas em amostragem para calcular densidades marginais . Journal of the American Statistical Association.
Outra maneira pela qual a computação rápida revolucionou as estatasticas e o aprendizado de ma¡quina épor meio de modelos bayesianos abertos. Os modelos estatasticos tradicionais são esta¡ticos: ajuste a distribuição A aos dados do tipo B. Mas a modelagem estatastica moderna tem uma qualidade mais Tinkertoy que permite resolver problemas com flexibilidade a medida que surgem, recorrendo a bibliotecas de distribuições e transformações. Precisamos apenas de ferramentas computacionais para ajustar esses modelos ajustados. Em seu influente artigo, Gelfand e Smith não desenvolveram nenhuma ferramenta nova; eles demonstraram como a amostragem de Gibbs pode ser usada para ajustar uma grande classe de modelos estatasticos. Nas últimas décadas, o amostrador de Gibbs foi substituado pelo Hamiltoniano Monte Carlo, filtragem departículas, Bayes variacional e algoritmos mais elaborados, mas o princapio geral de construção de modelo modular permaneceu.
6. Guido Imbens e Joshua Angrist (1994). Identificação e estimativa dos efeitos manãdios locais do tratamento. Econometrica.
A inferaªncia causal écentral para qualquer problema em que a pergunta não seja apenas uma descrição (como foram as coisas?) Ou previsão (o que acontecera¡ a seguir?), Mas um contrafactual (se fizermos X, o que acontecera¡ com Y?) . Os manãtodos causais evoluaram com o resto da estatastica e do aprendizado de ma¡quina por meio da exploração, modelagem e computação. Mas o raciocanio causal tem o desafio adicional de perguntar sobre dados que são impossaveis de medir (vocênão pode fazer X e não-X para a mesma pessoa). Como resultado, uma ideia chave neste campo éidentificar quais perguntas podem ser respondidas de forma confia¡vel a partir de um determinado experimento. Imbens e Angrist são economistas que escreveram um artigo influente sobre o que pode ser estimado quando os efeitos causais variam, e suas ideias formam a base para muitos dos trabalhos posteriores sobre este ta³pico.
7. Robert Tibshirani (1996). Redução da regressão e seleção por meio do laa§o . Journal of the Royal Statistical Society.
Na regressão, ou na previsão de uma varia¡vel de resultado a partir de um conjunto de entradas ou recursos, o desafio estãoem incluir muitas entradas junto com suas interações; o problema de estimativa resultante torna-se estatisticamente insta¡vel devido a s muitas maneiras diferentes de combinar essas entradas para obter previsaµes razoa¡veis. Manimos quadrados cla¡ssicos ou estimativas de máxima verossimilhana§a sera£o ruidosas e podem não funcionar bem em dados futuros e, portanto, vários manãtodos foram desenvolvidos para restringir ou “regularizar†o ajuste para obter estabilidade. Neste artigo, Tibshirani introduziu o lasso, uma abordagem computacionalmente eficiente e agora amplamente usada para regularização, que se tornou um modelo para regularização baseada em dados em modelos mais complicados.
8. Leland Wilkinson (1999). A grama¡tica dos gra¡ficos .
Neste livro, Wilkinson, um estatastico que trabalhou em vários projetos de software comercial influentes, incluindo SPSS e Tableau, apresenta uma estrutura para gra¡ficos estatasticos que vai além do foco usual em gra¡ficos de pizza versus histogramas, como desenhar um gra¡fico de dispersão e dados de tinta e chartjunk, para explorar abstratamente como os dados e as visualizações se relacionam. Este trabalho influenciou a estatastica por meio de muitos caminhos, principalmente por meio do ggplot2 e da familia de pacotes tidyverse na linguagem de computação R. a‰ um passo importante para integrar dados explorata³rios e análise de modelo no fluxo de trabalho da ciência de dados.
9. Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville e Yoshua Bengio (2014). Redes Adversariais Generativas . Proceedings of the International Conference on Neural Information Processing Systems.
Uma das conquistas impressionantes do aprendizado de ma¡quina nos últimos anos éa tomada de decisaµes em tempo real por meio de previsaµes e feedbacks de inferaªncia. Exemplos famosos incluem carros auta´nomos e AlphaGo da DeepMind, que se treinou para se tornar o melhor jogador de Go do planeta. Redes adversa¡rias gerativas, ou GANs, são um avanço conceitual que permite que problemas de aprendizagem por reforço sejam resolvidos automaticamente. Eles marcam um passo em direção ao objetivo de longa data da inteligaªncia artificial geral, ao mesmo tempo que aproveitam o poder do processamento paralelo para que um programa possa se treinar jogando milhões de jogos contra si mesmo. Em umnívelconceitual, os GANs vinculam a predição aos modelos generativos.
10. Yoshua Bengio, Yann LeCun e Geoffrey Hinton (2015). Aprendizado profundo . Nature .
Aprendizado profundo éuma classe de modelos de redes neurais artificiais que podem ser usados ​​para fazer previsaµes não lineares flexaveis usando um grande número de recursos. Seus blocos de construção - regressão logastica, estrutura de vários naveis e inferaªncia bayesiana - dificilmente são novos. O que torna essa linha de pesquisa tão influente éo reconhecimento de que esses modelos podem ser ajustados para resolver uma variedade de problemas de previsão, desde o comportamento do consumidor atéa análise de imagens. Como acontece com outros desenvolvimentos em estatastica e aprendizado de ma¡quina, o processo de ajuste são foi possível com o advento da computação paralela rápida e algoritmos estatasticos para aproveitar esse poder para ajustar modelos grandes em tempo real. Conceitualmente, ainda estamos alcana§ando o poder desses manãtodos, e épor isso que hátanto interesse em aprendizado de ma¡quina interpreta¡vel.