DrivenData blog

Learn how these top teams built the most accurate ways to privatize millions of taxi rides.

We'll show you how to get started with the residual U-NET benchmark model, a state-of-the-art approach to predicting the height and pose of ground objects for monocular satellite images taken from oblique angles.

Meet the winners who were best able to predict disturbances in Earth's magnetic field, and hear them explain their creative solutions.

Learn how these top teams built the most accurate ways to privatize census records.

Meet the winners who were best able to estimate the wind speeds of tropical storms at different points in time using satellite images.

Check out the winning solutions to identify the lab-of-origin for genetically engineered DNA, and meet their leaderboard-topping creators.

Hear from the winners who were best able to detect hate speech in memes using multi-modal models.

Hear how these top teams built the most accurate ways to privatize emergency call data in Baltimore.

We'll show you how to train an LSTM network using space weather data to predict changes in Earth's magnetic field for our latest challenge.

We'll show you how to start predicting wind speeds for our latest disaster preparedness challenge.

Hear from the winners who were best able to diagnose whole slide images of cervical biopsies using computer vision.

Learn how DrivenData uses nbautoexport, an open-source tool we developed, to make Jupyter Notebook code reviews easier.

The Community Spotlight features fantastic members from our DrivenData community. Quy Nguyen, a data scientist from Vietnam, talks about building time series models with lag features and the importance of storytelling in data science.

The Community Spotlight features fantastic members from our DrivenData community. Will Koehrsen, a full-stack data scientist from central Illinois, discusses tackling climate change with data science, being self-taught, and his favorite data science tools.

There's a lot at stake in the 2020 U.S. election. Can you predict the outcome? This benchmark shows you how to try using pollster data from FiveThirtyEight.

Meet the winners who were best able to detect clogged blood vessels and advance Alzheimer's research from 3D image stacks.

In this post we'll show you how to train a random forest model to predict the lab-of-origin from DNA sequences!

The Community Spotlight features fantastic members from our DrivenData community. Cecil Rivers, an electrical engineer in West Hartford, USA, talks about tricks for image processing, recruiting for diversity in the sciences, and his favorite sci-fi reads.

The Community Spotlight features fantastic members from our DrivenData community. Lamia Zaghloul, a data analysis consultant in Marseille, France, discusses fog nets, generative art, and using data science competitions to "study up" the rich and powerful.

Use what you've learned about AzureML to create and run your first machine learning pipeline in the AzureML cloud.