Methods and systems for high confidence utilization of datasets
View Patent ↗Methods and systems for high-confidence utilization of datasets are disclosed. In one embodiment, the method includes selecting a metric for determining substantially optimal combination of true positives and false positives in a data set, applying an optimization technique, and obtaining, from the results of the optimization technique, a value for at least one optimization parameter, the value for at least one optimization parameter resulting in substantially optimal combination of true positives and false positives. A number of true positives and a number of false positives are a function of the one or more optimization parameters.
1. A method for normalizing data from a target data set, the method being implemented in a computer and comprising the steps of:
sorting data from a reference data set according to measurement value;
selecting, from the sorted data, according to a predetermined criterion, reference subsets, the reference subsets having at least one reference measurement value;
selecting, utilizing the reference subsets, data elements in the target data set substantially equivalent to one reference subset;
sorting the data elements substantially equivalent to the one reference subset by measurement values, the sorted data elements comprising a sorted substantially equivalent subset; and
normalizing the target data set, utilizing the at least one reference measurement value and the sorted substantially equivalent subset;
wherein the steps of selecting data elements in said target data set, sorting the data elements and normalizing said target data set being performed by means of a non-transitory computer usable medium having computer readable code that causes the computer to perform the steps;
wherein the method is utilized for decision making to increase confidence on the use of the data in activities including manufacturing, handling, hybridization, and gene expression.
2. The method of claim 1 wherein the step of utilizing the at least one reference measurement value on the sorted substantially equivalent subset to normalize the data set comprises the step of utilizing a predetermined combination of the at least one reference value and the sorted substantially equivalent subset in order to establish a distribution curve.
3. The method of claim 2 wherein the step of utilizing a predetermined combination of the at least one reference value and the sorted substantially equivalent subset comprises the steps of:
iterating over portions of the sorted substantially equivalent subset in order to obtain the predetermined combination of the at least one reference value and the sorted substantially equivalent subset.
4. A computer program product comprising g non-transitory, computer usable medium having computer readable code embodied therein; said computer readable code being capable of causing a computer system to:
sort data from a reference data set according to measurement value;
select, from the sorted data, according to a predetermined criterion, reference subsets, the reference subsets having at least one reference measurement value;
select, utilizing the reference subsets, data elements in a target data set, the data elements in said target data set being substantially equivalent to one reference subset;
sort the data elements substantially equivalent to the one reference subset by measurement values, the sorted data elements comprising a sorted substantially equivalent subset; and
normalize said target data set, utilizing the at least one reference measurement value and the sorted substantially equivalent subset;
wherein normalize said target data set is utilized for decision making to increase confidence on the use of the data in activities including manufacturing, handling, hybridization and gene expression.