Where Can You Find Datasets for Deep Learning? | 2020

TL;DR

This video provides a quick overview of the top three platforms where you can find structured data for machine learning projects.

Transcript

what's up data Wranglers welcome back to neural net dot AI I am your host Phil Taber so I was talking to subscriber about where to find datasets and it occurred to me that there really isn't a good comprehensive list out there that's easy to find so I thought I'd go ahead and make a video really quick number one is going to be Kaggle so Kaggle is a... Read More

Key Insights

• 🥺 Kaggle is a leading platform for accessing a wide range of structured data sets for machine learning projects, offering competitions and prizes.
• 🎨 Despite its outdated design, the UCI Machine Learning Repository remains a reliable source for structured data, regularly updating its collection.
• 😫 Google's Data Set Search provides a convenient way to find specific data sets for machine learning projects, catering to various needs and interests.
• ℹ️ Other recommended data sources include data.gov, Datahub.io, and OpenML, offering diverse and collaborative options for structured data.
• 🎰 Accessing structured data is crucial for successful machine learning, deep learning, and data science projects.
• 😫 The availability of specific data sets, such as wine quality and census data, enables researchers to tackle different industry-related problems.
• 💨 Google's Data Set Search is a beta feature that should be utilized while available, providing a unique and user-friendly way to find structured data.

Questions & Answers

Q: What is Kaggle and why is it a great source for structured data?

Kaggle is a platform where companies host competitions and provide data sets for participants. It offers a diverse range of structured data, making it a valuable resource for machine learning projects. With data sets like wine reviews and census data, Kaggle caters to various industries and interests.

Q: Is the UCI Machine Learning Repository still reliable despite its outdated design?

Yes, the UCI Machine Learning Repository may have an outdated interface, but it remains a reliable source for structured data. The repository is regularly updated with new data sets and offers unique collections like superconductivity data, ensuring its relevance for data science projects.

Q: How can Google's Data Set Search benefit machine learning projects?

Google's Data Set Search allows users to find specific data sets for machine learning projects. By typing in keywords, like "handwritten," researchers can quickly access relevant data sets. Although still in beta, the search feature provides a helpful resource for finding relevant structured data.

Q: Are there any other recommended data sources for machine learning projects?

Apart from the mentioned platforms, there are several other data sources worth exploring. Examples include data.gov for US government data, Datahub.io for a wide array of data sets, and OpenML for collaborative machine learning projects. These platforms offer diverse and reliable sources for structured data.

Summary & Key Takeaways

• Kaggle: A popular website hosting competitions and providing a wide range of structured data sets, including wine reviews, census data, and taxi trajectory data.

• UCI Machine Learning Repository: Despite its outdated design, this repository offers regularly updated data sets, such as superconductivity data, making it a reliable source for structured data.

• Google Data Set Search: Google's beta feature allows users to find specific data sets for machine learning projects, such as handwritten data, making it a valuable resource.