Por que métricas importam em Machine Learning? | Métricas para Machine Learning #01 episode artwork

EPISODE · Jun 8, 2026 · 19 MIN

Por que métricas importam em Machine Learning? | Métricas para Machine Learning #01

from Prof. Rafael Kunst

Você já viu alguém afirmar que um modelo de Machine Learning é excelente porque atingiu 99% de acurácia?Neste primeiro episódio da série Métricas para Machine Learning, discutimos por que essa conclusão pode estar completamente errada.A avaliação de modelos é uma das etapas mais importantes em qualquer projeto de Inteligência Artificial. No entanto, muitas vezes a atenção fica concentrada nos algoritmos, enquanto a escolha das métricas recebe menos atenção do que deveria.Ao longo deste episódio, exploramos o papel das métricas na avaliação de modelos, analisamos um exemplo clássico envolvendo diagnóstico médico, discutimos o impacto das classes desbalanceadas e mostramos por que diferentes aplicações exigem diferentes formas de avaliação.Também apresentamos uma visão geral das principais métricas que serão abordadas ao longo da série, incluindo Accuracy, Precision, Recall, F1-Score e ROC/AUC.Neste episódioPor que métricas importamLimitações da acuráciaClasses balanceadas e desbalanceadasO custo dos erros em diferentes aplicaçõesAccuracy, Precision, Recall, F1 e ROC/AUCIntrodução à Matriz de ConfusãoSérie: Métricas para Machine LearningNesta série você aprenderá:Matriz de ConfusãoAcurácia (Accuracy)Precisão (Precision)Revocação (Recall)F1-ScoreROC e AUCMétricas para RegressãoComo escolher a métrica corretaPorque em Machine Learning não existe a melhor métrica. Existe a métrica adequada para o seu problema.

NOW PLAYING

Por que métricas importam em Machine Learning? | Métricas para Machine Learning #01

0:00 19:50

No transcript for this episode yet

We transcribe on demand. Request one and we'll notify you when it's ready — usually under 10 minutes.

What Works? Sophie Scott, UCL PALS Prof Sophie Scott, Director of the Institute of Cognitive Neuroscience at University College London, discusses life and science and careers with her colleagues from the Division of Psychology and Language Sciences at UCL, and beyond. The aim of the show is to highlight some amazing scientists, and explore their journeys through science and life, and find out what works for them. Block52 - Blockchain, Crypto Assets and DLT Prof. Dr. Philipp Sandner, Katharina Gehra, Philipp Schulden Weekly News about Blockchain, Crypto Assets and DLT. WszystkoWszedzie Rafael Markievitz Codzienne ciekawe 15 min fakty z wielu różnych dziedzin, podcast dla głodnych umysłów. Codziennie inny fascynujący temat: geografia, historia, matematyka, gospodarka, sport i inne.Najlepsze do słuchania w podróży, przy gotowaniu, sprzątaniu - spraw aby Twój czas był  spędzony w sposób wartościowy i inspirujący, nawet gdy robisz coś banalnego i nudnego jak stanie w korku czy czekanie na autobus, zapraszam do słuchania.Podcast był inspirowany podcastem Garego Arndt Everything Everywhere Daily - polecam serdecznie,Rafal Markievitz#wszystkowszedzie #wszystkowszędzie #podcastkontakt: [email protected]: https://www.instagram.com/wszystkowszedziepl/Facebook: https://www.facebook.com/wszystkowszedziepl Twitter X: <a href="https://www.youtube.com/redirect?event=channel_description&redir_token=QUFFLUhqazZPSkdOc19UODNZVWdBWlpvR1hjMmJTY2NiQXxBQ3Jtc0tsWDNLN1B5VnlBSEhjUVV6VXpaSGlWeWV0 The Injury Prevention Academy Podcast DORN Companies Welcome to The Injury Prevention Academy Podcast with DORN!Tune in for your ultimate source of cutting-edge insights on workplace injury prevention, safety, ergonomics and wellness. Hosted by DORN and Cheryl Roy, this podcast is your go-to destination for staying informed about the latest news, trends, and data in the realm of employee well-being and workplace safety.Join us as we bring you expert interviews and thought-provoking discussions with leading voices in the field. Our goal? Empowering you to create safer, healthier work environments for your valued employees.🌟 Key Highlights 🌟🔍 Stay Updated: Get the freshest news and data surrounding workplace injury prevention, ergonomics and safety.🧠 Expert Insights: Discover valuable insights from experts covering pain management, injury prevention, safety programs and technology.🤝 Supportive Strategies: Gain actionable strategies to prioritize the safety and well-being of your employees.Whether you're a business owner, HR prof

Frequently Asked Questions

How long is this episode of Prof. Rafael Kunst?

This episode is 19 minutes long.

When was this Prof. Rafael Kunst episode published?

This episode was published on June 8, 2026.

What is this episode about?

Você já viu alguém afirmar que um modelo de Machine Learning é excelente porque atingiu 99% de acurácia?Neste primeiro episódio da série Métricas para Machine Learning, discutimos por que essa conclusão pode estar completamente errada.A avaliação de...

Can I download this Prof. Rafael Kunst episode?

Yes, you can download this episode by clicking the download button on the episode player, or subscribe to the podcast in your preferred podcast app for automatic downloads.
URL copied to clipboard!