I want to collect a lot of data on my raspberry pi server for playing with basic neural networks. Any ideas for data source?

Question

I am following google developers YouTube channel and found AI really interesting. I also see neural networks being used in a lot of companies for all kinds of purposes. I want to experiment for instance with data from webscraping or with robot parts as sensors etc. Let me hear your ideas! :-)

Answer 1

iirc, NASA has lots of data for free. also both the US (USGS) and Australian geological services have free data for use. now if you want text, in the past, I have downloaded text from project Gutenberg to test my hadoop and spark programs. you might want to try that as well. finally, you can always generate random values or synthetic data. it just depends on what you want to do.

in the meantime, this should get you started:

NASA open data - https://open.nasa.gov/open-data/
USGS seismic marine data - https://walrus.wr.usgs.gov/NAMSS/
Australian seismic data sets - https://data.gov.au/dataset?tags=seismics
Project Gutenberg - https://www.gutenberg.org

-M

PS,

I just read the rest of your post. if you're interested, Magi Pi had a few projects in the 2016 big book of projects where Pis were reading sensors. one I recall was reading a thermometer. if was fairly simple to connect the sensor to the Pi(s). the biggest headache was converting digital to analog. that might be a simple way to get data into your Pi.