Indexing of large scale patient set

• US 9,305,039 B2
• Filed: 08/14/2013
• Issued: 04/05/2016
• Est. Priority Date: 12/19/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Wang, Jun, Wang, Fei
• Primary Examiner(s): Vital, Pierre
• Assistant Examiner(s): Sultana, Nargis
• Application Number:
  US13/967,062
• Publication Number:
  US 20140172870A1
• Time in Patent Office: 965 Days
• Field of Search: 707/741
• US Class Current: 1/1
• CPC Class Codes: G06F 16/13 File access structures, e.g. distributed indices arrangements of input from, or output to, record carriers G06F3/06
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/2246 Trees, e.g. B+trees
• CPC Class Codes: G06F 16/278 Data partitioning, e.g. horizontal or vertical partitioning
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/31 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/316 Indexing structures
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

Systems and methods for indexing data include formulating an objective function to index a dataset, a portion of the dataset including supervision information. A data property component of the objective function is determined, which utilizes a property of the dataset to group data of the dataset. A supervised component of the objective function is determined, which utilizes the supervision information to group data of the dataset. The objective function is optimized using a processor based upon the data property component and the supervised component to partition anode into a plurality of child nodes.

Claims

1. A non-transitory computer readable storage medium comprising a computer readable program for indexing data, wherein the computer readable program when executed on a computer causes the computer to perform the steps of:

formulating an objective function to index a dataset, a portion of the dataset including supervision information;

determining a data property component of the objective function which utilizes a property of the dataset to group data of the dataset, wherein determining the data property component includes maximizing a variance of the dataset, the maximizing the variance including applying maximum variance unfolding by forming a projection as an inner product of a data matrix and a learned partition hyperplane and maximizing overall pairwise distances in a projected space;

determining a supervised component of the objective function which utilizes the supervision information to group data of the dataset; and

optimizing the objective function using a processor based upon the data property component and the supervised component to partition a node into a plurality of child nodes.

2. The system as recited in claim 1, wherein the variance maximization module is further configured to apply a principal component analysis.

3. The non-transitory computer readable storage medium as recited in claim 1, wherein optimizing the objective function includes recursively optimizing the objective function at each node to provide a binary space partitioning tree.

4. The non-transitory computer readable storage medium as recited in claim 1, wherein maximizing the variance further includes applying principal component analysis.

5. The non-transitory computer readable storage medium as recited in claim 1, wherein determining the data property component includes clustering data around a median of the dataset.

6. The non-transitory computer readable storage medium as recited in claim 1, wherein determining the supervised component includes minimizing pairwise distance of similar data identified by the supervision information and maximizing pairwise distance of dissimilar data identified by the supervision information.

7. The non-transitory computer readable storage medium as recited in claim 1, wherein determining the supervised component includes applying the dataset as a linear prediction function.

8. The non-transitory computer readable storage medium as recited in claim 1, wherein the objective function includes a tradeoff parameter to balance contributions from the data property component and the supervision component.

9. The non-transitory computer readable storage medium as recited in claim 1, wherein the supervision information includes must-links indicating pairs of data that are similar to each other and cannot-links indicating pairs of data that are dissimilar to each other.

10. The non-transitory computer readable storage medium as recited in claim 1, wherein the dataset includes patient medical information.

11. A system for indexing data, comprising:

a formulation module configured to formulate an objective function to index a dataset stored on a non-transitory computer readable storage medium, a portion of the dataset including supervision information;

a data property module configured to determine a data property component of the objective function which utilizes a property of the dataset to group data of the dataset, the data property module including a variance maximization module configured to maximize a variance of the dataset, the maximizing the variance including applying maximum variance unfolding by forming a projection as an inner product of a data matrix and a learned partition hyperplane and maximizing overall pairwise distances in a projected space;

a supervision module configured to determine a supervised component of the objective function which utilizes the supervision information to group data of the dataset; and

an optimization module configured to optimize the objective function based upon the data property component and the supervised component to partition a node into a plurality of child nodes.

12. The system as recited in claim 11, wherein the optimization module is further configured to recursively optimize the objective function at each node to provide a binary space partitioning tree.

13. The system as recited in claim 11, wherein the data property module further comprises a cluster separation module configured to cluster data around a median of the dataset.

14. The system as recited in claim 11, wherein the supervision module further comprises a projection perspective module configured to minimize pairwise distance of similar data identified by the supervision information and maximize pairwise distance of dissimilar data identified by the supervision information.

15. The system as recited in claim 11, wherein the supervision module further comprises a prediction perspective module configured to apply the dataset as a linear prediction function.

16. The system as recited in claim 11, wherein the objective function includes a tradeoff parameter to balance contributions from the data property component and the supervision component.

17. The system as recited in claim 11, wherein the supervision information includes must-links indicating pairs of data that are similar to each other and cannot-links indicating pairs of data that are dissimilar to each other.