Croptimizer

• 

  This heatmap shows correlations in agri-data (N, P, K, temp, humidity). Red = strong positive, blue = negative. Useful for crop optimization

• 

  Croptimizer helps optimize crop selection based on location, soil type, and preferences for efficient and sustainable agriculture.

• 

• 

• 

  SHAP plot shows P, humidity, K, N as key features. Red = high, blue = low. SHAP values indicate each feature’s impact on model predictions.

Project Inspiration

Our project was inspired by a conversation with a farmer who wanted a free, all-in-one tool to predict crop yield and recommend the best crops based only on location-derived factors. They emphasized the need for a non-invasive solution that doesn’t require excessive data input.

Identifying the Problem

After researching, we found that many farmers face the same problem. Most existing tools either require too much personal data or focus on just one aspect, like soil nutrients or weather. We wanted to create a simple, effective solution.

Solution: Croptimizer

We built Croptimizer using machine learning to analyze multiple environmental factors from location data, predicting yields and recommending the most suitable crops.

Challenges Faced

• Ensuring accurate predictions with limited user input while keeping the tool easy to use.
• Balancing usability and accuracy, tweaking our model to improve results without overcomplicating the process.

Key Learnings

Through this project, we learned a lot about agriculture, AI, ML, data science, and user-friendly design. We hope Croptimizer helps farmers make informed decisions and improve their yields effortlessly!

Dataset Sources

• https://www.kaggle.com/datasets/datasetengineer/smart-farming-data-2024-sf24?resource=download

• https://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx

• https://nutrien-ekonomics.com/geo-data-analysis/

Built With

• api
• blinker
• certifi
• charset-normalizer
• click
• cloudpickle
• contourpy
• css3
• cycler
• et-xmlfile
• filelock
• flask
• flask-cors
• fonttools
• fsspec
• google
• html5
• idna
• itsdangerous
• jinja
• joblib
• kaggle
• kiwisolver
• llvmlite
• maps
• markupsafe
• matplotlib
• mpmath
• networkx
• nextjs
• numba
• numpy
• open-meteo
• openpyxl
• packaging
• pandas
• pillow
• pyparsing
• python
• python-dateutil
• pytz
• requests
• scikit-learn
• scipy
• seaborn
• setuptools
• shap
• six
• slicer
• sympy
• threadpoolctl
• torch
• tqdm
• typing-extensions
• tzdata
• urllib3
• werkzeug