Skip to content

Latest commit

 

History

History
71 lines (41 loc) · 1.95 KB

README.md

File metadata and controls

71 lines (41 loc) · 1.95 KB

ML based Crop Recommendation Appliccation

Technical Description

This app uses a Dataset with 2200 Datapoins for Training and for Generating predictions.
Model used: Gaussian Naive Bayes
Accuracy: 99.54%

Run it on your Browser now! (New)

How to run on Local Machine

  • Download the Github Package from this repo and Unzip it anywhere.

  • Download and install Anaconda for Windows from this link or Jupyter for Windows from this link.

  • Open Jupyter Notebook and navigate to the Crop-Recommendation folder.

  • Launch a new Jupyter Terminal and type these commands

 pip install streamlit
 pip install pandas
 pip install scikit-learn

  • Now navigate to Crop-Recommendation folder using cd command in the terminal.

  • Type this command

    streamlit run app.py
  • Enjoy!

Inspiration

Many farmers are confused when making the choice before the sowing season. This app will help them with their choice and save them a lot of time and money.

What it does

It takes input about the Farmer's soil and tells them which Crop would be best for their soil type using ML prediction.

How we built it

We have used a dataset of 2200 entries and trained an ML model on it for making the predictions. We have then used Streamlit library to create a user-friendly and simple UI for anyone to use. It uses Jupyter environment to run.

Challenges we ran into

Finding the right dataset, Training the model for high accuracy, Hosting the model as a web-app.

Accomplishments that we're proud of

High Accuracy of our ML model (99.54% during validation). Custom UI elements on Streamlit web-app.

What we learned

Machine Learning, Using Streamlit for Hosting, Creating custom environment on Jupyter.