Automated account opening decisioning using machine learning

1. A method for automatically opening an account through a received account opening request using machine learning performed on circuitry, the method comprising:

creating a plurality of machine learning rules engines, using the circuitry, by accessing past decisions made regarding past account opening requests stored as past account opening records in a memory comprising non-transitory computer readable media,

the plurality of machine learning rules engines created by executing a machine learning algorithm on each subset of the past account opening records,

where each machine learning algorithm creates a quality control measure comprising an F-score that is combined with F-Scores from each of the machine learning algorithms to determine an optimal set of machine learning rules for the plurality of machine learning rules engines that are generated by ranking the F-score for each rule and selecting rules with a highest F-score as compared to other F-scores,

wherein the F-score is calculated based on a precision value comprising a number of correct results divided by a number of all returned results and a recall value comprising the number of correct results divided by a number of results that should have been returned,

wherein the plurality of machine learning rules engines include a grant rules engine,

wherein the past account opening records each include a result comprising a grant or denial of the past account opening requests associated with the past account opening records, and

properties of the past account opening requests associated with the past account opening records including a risk score determined for the past account opening requests associated with the past account opening records,

wherein the past account opening records used to train the grant rules engine include only granted account opening records;

receiving, with the circuitry, the received account opening request;

determining a recommendation, using the circuitry, for granting or denying the received account opening request, wherein the determination comprises executing the grant rules engine on properties of the received account opening request using the circuitry, thus creating a grant rules engine result; and

granting or denying the received account opening request, using the circuitry, based on the grant rules engine result compared to a predetermined grant threshold.

2. The method of claim 1 , wherein the plurality of machine learning rules engines also includes a denial rules engine.

3. The method of claim 2 , wherein the determining of the recommendation also comprises executing the denial rules engine on the properties of the received account opening request using the circuitry, thus creating a denial rules engine result, wherein the past account opening records used to create the denial rules engine include only denied account opening records.

4. The method of claim 3 , wherein the determining of the recommendation further comprises mathematically combining the grant rules engine result with the denial rules engine result.

5. The method of claim 4 , wherein the granting or denying of the received account opening request is based on a comparison of the mathematical combination of the grant rules engine result with the denial rules engine result with a threshold.

6. The method of claim 1 , wherein the granting or denying of the received account opening request is based on a comparison of the grant rules engine result with a threshold.

7. The method of claim 1 , wherein the plurality of machine learning rules engines also includes a cross sales rules engine.

8. The method of claim 7 , further comprising prompting a user, using the circuitry, for additional products using a second customized screen generated by the cross sales rules engine where the cross sales rules engine determines a second selection and a second order of cross sales products on the second customized screen.

9. The method of claim 1 , wherein the machine learning algorithm is a Densicube algorithm.

10. The method of claim 1 , wherein the machine learning algorithm is a K-Means algorithm.

11. The method of claim 1 , wherein the machine learning algorithm is a Random Forest algorithm.

12. The method of claim 1 , wherein the plurality of machine learning rules engines are created in parallel.

13. The method of claim 1 , wherein the account is a checking account.

14. The method of claim 1 , wherein the account is a loan account.

15. A computing device for automatically opening an account through a received account opening request using machine learning, the computing device comprising:

memory comprising non-transitory computer readable media, wherein the memory stores past decisions made regarding past account opening requests as past account opening records,

the past account opening records each include a result comprising grant or denial of the past account opening requests associated with the past account opening records, and past properties of the past account opening requests associated with the past account opening records including risk scores determined for the past account opening requests associated with the past account opening records;

the memory also stores a plurality of machine learning rules engines, the plurality of machine learning rules engines created by executing a machine learning algorithm on each subset of the past account opening records,

where each machine learning algorithm creates quality control measure comprising an F-score that is combined with F-scores from each of the machine learning algorithms to determine an optimal set of machine learning rules for the plurality of machine learning rules engines that are generated by ranking the F-score for each rule and selecting rules with a highest F-score as compared to other F-scores,

wherein the F-score is calculated based on a precision value comprising a number of correct results divided by a number of all returned results and a recall value comprising the number of correct results divided by a number of results that should have been returned,

wherein the plurality of machine learning rules engines include a grant rules engine, wherein the past account opening records used to train the grant rules engine include only granted account opening records;

the memory also stores the received account opening request which includes received properties, the received properties including a risk score for the received account opening request; and circuitry configured to:

receive the received account opening request;

determine a recommendation for granting or denying the received account opening request, wherein the determination comprises execution of the grant rules engine on the received properties of the received account opening request, the execution creating a grant rules engine result; and

automatically grant or deny the received account opening request based on the grant rules engine result compared to a predetermined grant threshold.

16. The computing device of claim 15 , wherein the received account opening request being reviewed includes missing data, inaccurate data, or an inconclusive risk score.

17. The computing device of claim 15 , wherein the received account opening request comprises at least one of a request to open the account at a financial institution or to add a service to the account.

18. The computing device of claim 15 , wherein the past properties of the past account opening records and the received properties of the received account opening request include at least one of a credit score, credit history, an annual income, occupation, debit tools, history of non-payment of accounts, past bankruptcy, investment portfolio, savings amount, or investment amount.

19. The computing device of claim 15 , wherein the account is a credit card account.