Predicting crop yield with a crop prediction engine

1. A method for predicting a crop yield for a particular type of seed in a particular location, the method comprising:

receiving, at one of a plurality of client devices, a plurality of seasonal crop data for the particular type of seed;

receiving, at one of the plurality of client devices, a plurality of soil data associated with the particular location, wherein the soil data includes a plurality of soil variables and the particular location is represented by a plurality of voxels;

receiving, at one of the plurality of client devices, a plurality of mapping data that can be associated with the particular location;

uploading the seasonal crop data, the soil data and the mapping data to a geospatial database associated with a crop prediction engine;

applying, at the crop prediction engine, a random forest prediction model to the seasonal crop data, the soil data and mapping data in the geospatial database;

ranking, at the crop prediction engine, a plurality of covariates to determine one or more significant covariates that are ranked higher than one or more remaining covariates of the plurality of covariates; and

re-applying, at the crop prediction engine, the one or more significant covariates to the random forest prediction model generates an estimated crop yield output for a specific voxel associated with the particular location and predicts the crop yield for the particular type of seed at the particular location.

2. The method of claim 1 wherein the ranking, at the crop prediction engine, of the plurality of covariates is performed with a random intersection tree (RIT) model.

3. The method of claim 1 further comprising applying, at the crop prediction engine, a validation process to determine an error for the predicted crop yield associated with the particular type of seed at the particular location.

4. The method of claim 1 further comprising predicting, at the crop prediction engine, another crop yield for a different type of seed at the same location.

5. The method of claim 1 further comprising receiving, at one of the plurality of clients, at least one or more nutrient data for the particular type of seed.

6. The method of claim 1 further comprising estimating, at the crop prediction engine, a crop growth to predict the crop yield.

7. The method of claim 1 further comprising determining, at the crop prediction engine, a crop cost based on the particular type of seed, a nutrient data for the particular type of seed and the crop yield.

8. The method of claim 1 wherein the output of the random forest prediction model generates a plurality of decision trees.

9. A method for predicting a crop yield for a particular type of seed in a particular location, the method comprising:

receiving, at one of a plurality of client devices, a plurality of seasonal crop data for the particular type of seed;

receiving, at one of the plurality of client devices, a plurality of soil data associated with the particular location, wherein the soil data includes a plurality of soil variables and the particular location is represented by a plurality of voxels;

receiving, at one of the plurality of client devices, a plurality of mapping data that can be associated with the particular location;

receiving, at one of the plurality of clients, at least one or more nutrient data for the particular type of seed;

uploading the seasonal crop data, the soil data and the mapping data to a geospatial database associated with a crop prediction engine;

applying, at the crop prediction engine, a random forest prediction model to the seasonal crop data, the soil data and mapping data in the geospatial database,

wherein the output of the random forest prediction model generates a plurality of decision trees;

ranking, at the crop prediction engine, a plurality of covariates to determine one or more significant covariates that are ranked higher than one or more remaining covariates of the plurality of covariates,

wherein the ranking of the plurality of covariates is performed with a random intersection tree (RIT) model;

re-applying, at the crop prediction engine, the one or more significant covariates to the random forest prediction model to predict the crop yield for the particular type of seed at the particular location; and

determining, at the crop prediction engine, a crop cost based on the particular type of seed, the nutrient data for the particular type of seed and the crop yield.

10. The method of claim 9 , wherein re-applying, at the crop prediction engine, the plurality of significant covariates to the random forest prediction model generates an estimated crop yield output for a specific voxel associated with the particular location.

11. The method of claim 10 further comprising predicting, at the crop prediction engine, another crop yield for a different type of seed at the same location.

12. The method of claim 10 further comprising applying, at the crop prediction engine, a validation process to determine an error for the predicted crop yield associated with the particular type of seed at the particular location.

13. The method of claim 12 further comprising estimating, at a crop prediction engine, a crop growth to predict the crop yield.

14. A system for predicting a crop yield for a particular type of seed in a particular location, the system comprising:

a client device that receives a plurality of seasonal crop data for the particular type of seed;

the client device receives a plurality of soil data associated with the particular location, wherein the soil data includes a plurality of soil variables and the particular location is represented by a plurality of voxels;

the client device receives a plurality of mapping data that can be associated with the particular location;

the client device receives at least one or more nutrient data for the particular type of seed;

a geospatial database associated with a crop prediction engine that is communicatively coupled to the client device,

wherein the geospatial database receives the seasonal crop data, the soil data and the mapping data;

the crop prediction engine applies a random forest prediction model to the seasonal crop data, the soil data and mapping data in the geospatial database,

wherein the output of the random forest prediction model generates a plurality of decision trees;

the crop prediction engine ranks a plurality of covariates to determine one or more significant covariates that are ranked higher than one or more remaining covariates of the plurality of covariates,

wherein the ranking of the plurality of covariates is performed with a random intersection tree (RIT) model; and

the crop prediction engine re-applies the one or more significant covariates to the random forest prediction model to predict the crop yield for the particular type of seed at the particular location; and

the crop prediction engine determines a crop cost based on the particular type of seed, a nutrient data for the particular type of seed and the crop yield.

15. The system of claim 14 wherein the crop prediction engine generates an estimated crop yield output for a specific voxel associated with the particular location.

16. The system of claim 14 wherein the crop prediction engine predicts another crop yield for a different type of seed at the same location.

17. The system of claim 14 wherein the crop prediction engine applies a validation process to determine an error for the predicted crop yield associated with the particular type of seed at the particular location.

18. The system of claim 14 wherein the crop prediction engine estimates a crop growth to predict the crop yield.