Learning data science is fun, so then why do we always use the same boring datasets? It's common to see projects using the iris, cars, or titanic data. Stand out! Check these 9 datasets on I created on #kaggle perfect for a unique portfolio project. #datascience#datasets🧵👇
1. MrBeast Youtube Stats
Includes metadata for every MrBeast Youtube video including: title, description, view, comment counts, likes AND thumbnails. Updated daily so you can track this viral sensation’s video trends over time.
Dataset of over 200k OSHA reportable injuries spanning 5 years. Do some investigative data science to see which industries produce the most injuries and which companies keep their employees safe.
This dataset contains metrics for ~1,000 different roller coasters from around the world. Includes tons of metrics like top speed, number of flips, year built, and even lat/lon locations so you can plot them on a map!
Want to beef up your computer vision skills? This is the perfect dataset with over ~1M high quality word annotations on images. Train a custom model capable of OCR text extraction.
This is the perfect dataset to learn binary classification. See if you can create an algorithm that uses EEG measurements to tell if the subject’s eyes are open or closed.
7. Monthly measurements of Zillow’s home values for each US state going back to 2000. This small dataset is perfect for any beginner interested in working with time series data.
🔗 kaggle.com/datasets/robik…
8. Annotated Car Driving Footage
Multi-Object tracking is one of the most cutting edge fields in computer vision. This dataset provides video footage of cars driving through cities and labels are provided for every car, pedestrian and stop light.
🔗 kaggle.com/datasets/robik…
9. Historic Global Exchange Rates
This dataset is updated daily with exchange rates from around the world. Do some data exploration to see if you can find unique trends related to world events.