Maximize the efficiency of fresh produce distribution is both important for farmers and customers. The work ItradeNetwork does inspire us to come up with a model that can predict the rejection of certain produce order. With the model, we can help reduce the current world issue about food waste.

What it does

Based on the exsiting dataset, we built and trained a failure prediction model. Given a set of features as inputs:

  • 'Category Name' -- name of the produce
    • (i.e. Apples)
  • 'Vendor Name' -- name of the Vendor prodiving that produce
    • (i.e. Vendor 1c47447e)
  • 'Shipping Warehouse' -- code refers to the Shipping Warehouse
    • (i.e Warehouse 60268bc9)
  • 'Inspector'-- name of the Inspector
    • (i.e Inspector afbf7591)
  • 'Shipping Time'-- duration of the shipment
    • (i.e 3)
  • 'Season' -- season of the shipment
    • (i.e Spring) Our model predicts whether more than 90% of the cases will be failed.

How we built it

We used Jupyter note with Python 3 to process our dataset given by ItradeNetwork. We cleaned the data, take a quick look at the characteristics of our data, created list of features to develop a Machine Learning model. We trained our model with 80% dataset we have and tested with the rest 20% of the dataset.

Challenges we ran into

  1. parsing out features that have reality applications
  2. process a have high dimension of categorical data.
  3. see how good is our model, find the best test
  4. linking the backend with the front end

Accomplishments that we're proud of

Our team pulled out each one's strength and we were very supportive to each other. We all learned a lot in this application of data science.

What we learned

  1. we learned a lot of ML tests for different goals.
  2. we learned how to use python better
  3. we learned how to link front and back end together.

What's next for HackXX with ItradeNetwork

  1. We can continue training the model to make our prediction better.
  2. Creating better webpage for user interation
  3. Develop a clear way of explanation of the model output

Built With

Share this project: