Methods and systems for performing correlation-based parking availability estimation

1. A method comprising:

retrieving a plurality of parking availability correlation models, wherein each of the parking availability correlation models estimates parking availability for a target parking block comprising a plurality of parking spaces based on parking availability for one or more other parking blocks;

determining a first subset of the one or more other parking blocks for which current parking availability is known;

selecting a second subset of the parking availability correlation models, wherein the second subset comprises parking availability correlation models that estimate parking availability of the target parking block based on parking availability for one or more parking blocks of the first subset of the one or more other parking blocks for which current parking availability is known;

selecting a parking availability correlation model from the second subset based on a model quality indicator associated with each parking availability correlation model; and

estimating parking availability for the target parking block based on the selected parking availability correlation model.

2. The method of claim 1 , wherein each of the parking availability correlation models estimates parking availability for the target parking block based on parking availability for a different combination of other parking blocks.

3. The method of claim 1 , wherein the parking availability for a parking block comprises a number of available parking spaces in the parking block.

4. The method of claim 1 , wherein the parking availability for a parking block comprises a percentage of a total number of parking spaces in the parking block that are available.

5. The method of claim 1 , further comprising:

determining the current parking availability for the one or more other parking blocks based on sensor data received from one or more connected vehicles located in the one or more other parking blocks.

6. The method of claim 1 , further comprising:

determining that current parking availability for a parking block is known when sensor data has been received from one or more vehicles located in the parking block within a threshold time period.

7. The method of claim 1 , wherein the model quality indicator associated with a parking availability correlation model comprises an indication as to how accurate the parking availability correlation model is likely to be.

8. The method of claim 1 , wherein each of the parking availability correlation models comprises a linear regression model.

9. The method of claim 1 , further comprising:

when the second subset of the parking availability correlation models comprises no parking availability correlation models, estimating the parking availability of the target parking block based on a default parking availability estimation model.

10. The method of claim 1 , further comprising:

training each of the parking availability correlation models based on historical parking availability data of the one or more other parking blocks and the target parking block.

11. The method of claim 10 , further comprising:

compiling training data comprising a plurality of training examples, wherein each training example comprises parking availability of the target parking block and parking availability of one or more other parking blocks at a certain time; and

training each of the parking availability correlation models using the training data.

12. The method of claim 11 , further comprising:

training each of the parking availability correlation models as a linear regression model.

13. The method of claim 12 , further comprising:

training each of the parking availability correlation models by selecting linear regression parameters such that a loss function is minimized, wherein the loss function is based on an error rate between the estimated parking availability for the target parking block and the actual parking availability for the target parking block.

14. The method of claim 13 , wherein the loss function comprises a mean square error.

15. The method of claim 13 , further comprising:

determining the model quality indicator for each parking availability correlation model based on the loss function associated with each parking availability correlation model.

16. A server comprising a controller configured to:

retrieve a plurality of parking availability correlation models, wherein each of the parking availability correlation models estimates parking availability for a target parking block comprising a plurality of parking spaces based on parking availability for one or more other parking blocks;

determine a first subset of the one or more other parking blocks for which current parking availability is known;

select a second subset of the parking availability correlation models, wherein the second subset comprises parking availability correlation models that estimate parking availability of the target parking block based on parking availability for one or more parking blocks of the first subset of the one or more other parking blocks for which current parking availability is known;

select a parking availability correlation model from the second subset based on a model quality indicator associated with each parking availability correlation model; and

estimate parking availability for the target parking block based on the selected parking availability correlation model.

17. The server of claim 16 , wherein the controller is further configured to:

determine the current parking availability for the one or more other parking blocks based on sensor data received from one or more connected vehicles located in the one or more other parking blocks; and

determine that current parking availability for a parking block is known when sensor data has been received from one or more vehicles located in the parking block within a threshold time period.

18. The server of claim 16 , wherein the controller is further configured to:

compile training data comprising a plurality of training examples, wherein each training example comprises parking availability of the target parking block and parking availability of one or more other parking blocks at a certain time; and

train each of the parking availability correlation model using the training data.

19. The server of claim 18 , wherein the controller is further configured to:

train each of the parking availability correlation models as a linear regression model by selecting linear regression parameters such that a loss function is minimized, wherein the loss function is based on an error rate between the estimated parking availability for the target parking block and the actual parking availability for the target parking block.

20. The server of claim 19 , wherein the controller is further configured to:

determine the model quality indicator for each parking availability correlation model based on the loss function associated with each parking availability correlation model.