Skip to content

9ooDa/Artificial-Ignorance_AI_CityHack22

Repository files navigation

Artificial-Ignorance_AI_CityHack22

Project: < Cabbage demand prediction >

Cabbage

Team: < Artificial Ignorance >

Members

  • < Dayeon Ku > (Leader)
  • < Minho Joh >
  • < Jerry Wu >
  • < Rahul >

Description of the Project

AI is one of the most powerful tools in the era of computer technology. With the help of AI, even traditional industries can benefit from it. In this project, we have made an AI predict the demand for vegetables such as cabbage. We all know that the price of vegetables can fluctuate a lot due to many factors, and this is precisely what our product features. Using this AI, we can predict the demand of our product for both long term and short term, and the time interval can be precise at every hour! With the help of this AI, farmers will no longer need to worry about producing too much or too little vegetables and stand a loss.
Also, this project helps solve the issue with the number of cabbages (vegetables) that we need to purchase from the farmers through the analysis (prediction) of the future demands for it.
We believe that this project can help both the farmers and the sellers in the market by predicting the future demands of the vegetables.
Benefits that can be obtained through this project are predictions for the future demand of vegetables, analysis and timely updates to the farmers, and the protection of the environment through the avoidance of wastage. We expect to gain more stable economic conditions through these benefits.
Some of the methods we used in the project were Random Forest Regressor, Linear Regression, ARIMA, ARIMAX, K-nearest neighbors (KNN), Decision Tree Classifier, Support Vector Regression (SVR), and Linear Discriminant Analysis (LDA). We found out that the Random Forest Regressor has the best R^2 value among all these algorithms.
In addition, as we sell our product to other sellers, we will be able to obtain new data for the number of demand and the features that can be later used to improve our prediction.

3 Most Impactful Features of the Project (with Screenshot and Short Description (150 words))

1. < Demand Prediction >


The key feature of our project is that it can predict the future demand of the cabbage through the given training data. This will allow higher profit and solve an environmental problem that initially occurred due to cabbage waste of unsold and remaining cabbages.

2. < Important features recognition >


Another essential feature of our project is recognizing the critical features for the prediction of demand. The two most important features that had an impact on the number of demand predictions were "Quality" and "Broker fee" (X1 and X2).

3. < Principal method and its hyperparameter selection >


The recognition and selection of the prediction method and its hyperparameters for our prediction is also one of the impactful feature of this Project. Trough the appliance of numerous different methods, we found that Random Forest Regressor returned the best R squared and prediction of the future demand. In addition, thorugh the modification of its parameters, we were able to find the best parameters (hyperparameter) for the accurate demand prediction.

Tech used (as many as required)

1. < Random Forest Regressor >

2. < Linear Regression >

3. < Support Vector Regression (SVR) >

4. < K-Nearest Neighbors (KNN) >

5. < Decision Tree Classifier >

6. < Linear Discriminant Analysis (LDA) >

7. < Gaussian NB >

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published