Inspiration
Every year, so much of land's potential goes to waste because of inaccurate measures and misinformation amongst many farmers. This not only affects the farmer but also the consumers worldwide, and hence the world's GDP. Growing up as an Indian seeing the issues faced by every other farmer inspired me to make a real solution that could help use our limited resources more efficiently.
What it does
Our web app takes in data like the nitrogenous content of the soil, phosphoric content of the soil, potassium content in the soil, humidity of the region, average temperature in the farming period, average rainfall received throughout the farming period, and the pH level of the soil. It then runs this data through our database of over 10,000+ successful farming produces using machine learning and predicts which crop best suits the soil and the region.
How we built it
We have used machine learning to find the maximum similarity between the farmer's (user) soil and the past records of previous soil samples present in our database. We then used Streamlit to host a webpage, and take the input from the user, and display the output based on our code's best predictions.
Challenges we ran into
We initially found the solution a little hard to come up with since none of us had used machine learning to develop such an app before, but we gradually learned how to use it with endless amounts of learning. We also found it really hard to make our Streamlit web app look prettier with styling because of so many restrictions in Streamlit, but eventually found our way through it.
Accomplishments that we're proud of
We successfully learned how to deploy machine learning to build this web app, and also learned how to use Streamlit from scratch. We had no prior experience with the both of them, but learning it and being able to come up with a really good product is the best accomplishment we could hope for.
What we learned
We learned how to use Streamlit and machine learning from scratch. Siddharth also got the opportunity to learn Github by sharing codes with Jonathan over there.
What's next for Crop Recommendations Predictor
We are looking forward to make an app that also starts giving notifications and alerts throughout the farming period for farmers to be able to bring out the 100% efficiency in farming.
Log in or sign up for Devpost to join the conversation.