[Feature Request]: Implement Explainable AI (XAI) for Crop Prediction Model

[kaishwarya24]

Is there an existing issue for this?

• [X] I have searched the existing issues

Feature Description

The current crop recommendation system employs various classification algorithms, with the Random Forest Classifier selected for deployment due to its high accuracy (99.45%). While the model demonstrates excellent predictive performance, it lacks transparency regarding how predictions are made.

To enhance the usability and trustworthiness of the model, I propose the integration of Explainable AI (XAI) techniques. This addition will allow users to understand the factors influencing the model’s predictions, which can help farmers make informed decisions based on the recommendations provided.

Use Case

Integration of XAI Techniques:
    Implement SHAP (SHapley Additive exPlanations) or LIME (Local Interpretable Model-agnostic Explanations) to provide insights into feature importance and model predictions.

Documentation:
    Update the documentation to include details on how to interpret the XAI outputs, ensuring users can easily understand the model’s decisions.

Visualization:
    Create visualizations (e.g., feature importance plots, individual prediction explanations) that can be integrated into the existing user interface, making the XAI outputs accessible and user-friendly.

Benefits

Enhanced Transparency: Users will gain insights into how specific soil and environmental conditions affect crop recommendations.
Informed Decision-Making: By understanding model predictions, farmers can better evaluate and trust the recommendations provided by the system.
Improved Model Validation: XAI will allow for more rigorous validation of the model, ensuring that it aligns with agronomic principles.

Add ScreenShots

No response

Priority

High

Record

• [X] I have read the Contributing Guidelines
• [X] I'm a GSSOC'24 contributor
• [X] I have starred the repository

[github-actions[bot]]

Thank you for creating this issue! 🎉 We'll look into it as soon as possible. In the meantime, please make sure to provide all the necessary details and context. If you have any questions or additional information, feel free to add them here. Your contributions are highly appreciated! 😊

You can also check our CONTRIBUTING.md for guidelines on contributing to this project.

[manikumarreddyu]

can you please let me know how much time do you need to complete this issue @kaishwarya24

[kaishwarya24]

@manikumarreddyu I will submit it by tomorrow As early as possible

[manikumarreddyu]

Any updates. @kaishwarya24

[kaishwarya24]

"Hi @manikumarreddyu, I’ve completed the work and am trying to upload the file, but it's too large (36.1 MB). I’ll upload it as soon as possible."

[manikumarreddyu]

can you please share the video..how it works

[kaishwarya24]

This is how they look. I’ve recorded the video, but I can’t upload it here since the file size is 15 MB. i will install git and upload it through command line, can i get some time

[kaishwarya24]

https://github.com/user-attachments/assets/29386fe0-4443-43a0-8dc1-35de3039be50

[manikumarreddyu]

hi @kaishwarya24 im sorry to inform you that jupyter notebooks are removed from the repo

[kaishwarya24]

yes @manikumarreddyu I have observed it and am about to ask you, then how should I proceed now?

[manikumarreddyu]

can you implement that feature in any other areas ..in website

[kaishwarya24]

I usually work a lot with ML and Deep learning But iam not much aware of deployment in frontend and backend etc.. but I can try if I have 2 days of time Or else I will upload this .ibynb and .csv in a folder

[manikumarreddyu]

no dont do that...i really appreciate your work...change any one line in readme.. i will merge it..so that you will get points..you really worked hard

[kaishwarya24]

Isn't it possible to merge this file without notebooks folder

[kaishwarya24]

Ok I'll do that Later on if any possible to add my work please message me

[github-actions[bot]]

Hello @kaishwarya24! Your issue #142 has been closed. Thank you for your contribution!