# US 2020/0003679 A1 · Cross-validation based calibration of a spectroscopic model

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0003679 A1
- Document ID: US20200003679A1
- Status: Published application
- Kind: A1
- Application number: 16/448,914
- Filed: 2019-06-21
- Published: 2020-01-02
- Earliest claimed: 2018-06-29
- Assignees: Viavi Solutions Inc.
- Applicants: VIAVI Solutions Inc.
- Inventors: Chang Meng HSIUNG; Lan SUN
- Web page: https://patentgrep.com/patent/US20200003679A1

## CPC

- G01N 21/274: {Calibration, base line adjustment, drift correction}
- G01J 3/0275: {making use of sensor-related data, e.g. for identification of sensor parts or optical elements}
- G01N 21/31: Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry {(G01N21/72 takes precedence)}
- G01N 2021/0118: {Apparatus with remote processing}
- G01N 2201/129: Using chemometrical methods
- G01N: INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES (measuring or testing apparatus or processes other than immunoassay, involving enzymes or microorganisms C12M, C12Q)
- G01J: MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY (light sources F21, H01J, H01K, H05B; investigating properties of materials by optical means G01N)

## Abstract

A device may receive a master data set for a first spectroscopic model; receive a target data set for a target population associated with the first spectroscopic model to update the first spectroscopic model; generate a training data set that includes the master data set and first data from the target data set; generate a validation data set that includes second data from the target data set and not the master data set; generate, using cross-validation and using the training data set and the validation data set, a second spectroscopic model that is an update of the first spectroscopic model; and provide the second spectroscopic model.

## Claims

### Claim 1

A device, comprising:
one or more memories; and
one or more processors communicatively coupled to the one or more memories, configured to:
receive a master data set for a first spectroscopic model;
receive a target data set for a target population associated with the first spectroscopic model to update the first spectroscopic model;
generate a training data set that includes the master data set and first data from the target data set;
generate a validation data set that includes second data from the target data set and not the master data set;
generate, using cross-validation and using the training data set and the validation data set, a second spectroscopic model that is an update of the first spectroscopic model; and
provide the second spectroscopic model.

### Claim 2 (depends on 1)

The device of claim 1, wherein the one or more processors are configured to:
receive a spectroscopic measurement;
perform a spectroscopic determination using the second spectroscopic model; and
provide an output identifying the spectroscopic determination.

### Claim 3 (depends on 1)

The device of claim 1, wherein the training data set is a plurality of training data sets and the validation data set is a plurality of validation data sets; and
wherein the one or more processors, when generating the second spectroscopic model, are configured to:
generate a plurality of performance metrics based on the plurality of training data sets and the plurality of validation data sets;
determine a total performance metric based on the plurality of performance metrics;
determine an optimal partial least squares (PLS) factor based on the total performance metric; and
determine the second spectroscopic model based on the optimal PLS factor and a merged data set,
wherein the merged data set includes the master data set and the target data set.

### Claim 4 (depends on 1)

The device of claim 1, wherein the first spectroscopic model and the second spectroscopic model are quantification models.

### Claim 5 (depends on 1)

The device of claim 1, wherein the master data set is based on a first set of spectroscopic measurements performed by a master spectrometer and the target data set is based on a second set of spectroscopic measurements performed by a target spectrometer that is different from the master spectrometer.

### Claim 6 (depends on 1)

The device of claim 1, wherein the master data set is based on a first set of spectroscopic measurements performed by a particular spectrometer and the target data set is based on a second set of spectroscopic measurements performed by the particular spectrometer.

### Claim 7

A method, comprising:
receiving, by a device, a target data set for a target population associated with a first spectroscopic model;
obtaining, by the device, a master data set for the first spectroscopic model based on receiving the target data set;
determining, by the device, an optimal partial least squares (PLS) factor using cross-validation,
wherein the optimal PLS factor is determined based on a plurality of training data sets, each training data set including a respective portion of the target data set and all of the master data set and based on a plurality of validation data sets, each validation data set including a respective portion of the target data set and not including data of the master data set;
merging, by the device, the target data set and the master data set to generate a merged data set;
generating, by the device and using the merged data set and the optimal PLS factor, a second spectroscopic model,
wherein the second spectroscopic model is an update of the first spectroscopic model; and
providing, by the device, the second spectroscopic model to replace the first spectroscopic model.

### Claim 8 (depends on 7)

The method of claim 7, wherein determining the optimal PLS factor comprises:
determining partial least squares (PLS) performance metrics for each of the plurality of training data sets and each of the plurality of validation data sets;
determining a total PLS performance metric based on the PLS performance metrics; and
optimizing a PLS factor to determine the optimal PLS factor for the second spectroscopic model based on the total PLS performance metric.

### Claim 9 (depends on 8)

The method of claim 8, wherein the total PLS performance metric is associated with a root mean square error (RMSE) value, and
wherein optimizing the PLS factor comprises:
optimizing the PLS factor to minimize the RMSE value.

### Claim 10 (depends on 8)

The method of claim 8, wherein the plurality of validation data sets includes different data of the target data set than the plurality of training data sets.

### Claim 11 (depends on 8)

The method of claim 8, wherein the determining the total PLS performance metric comprises:
aggregating the PLS performance metrics.

### Claim 12 (depends on 7)

The method of claim 7, wherein the target data set is associated with a set of measurements of the target population performed after measurements associated with the master data set.

### Claim 13 (depends on 12)

The method of claim 12, wherein the second spectroscopic model is a calibration update model of the first spectroscopic model.

### Claim 14 (depends on 7)

The method of claim 7, wherein the target data set is associated with a set of measurements performed by a particular spectrometer that is different from one or more spectrometers that performed measurements associated with the master data set.

### Claim 15 (depends on 14)

The method of claim 14, wherein the second spectroscopic model is a calibration transfer model of the first spectroscopic model.

### Claim 16 (depends on 14)

The method of claim 14, wherein providing the second spectroscopic model comprises:
providing the second spectroscopic model for use in connection with subsequent measurements by the particular spectrometer.

### Claim 17

A non-transitory computer-readable medium storing instructions, the instructions comprising:
one or more instructions that, when executed by one or more processors, cause the one or more processors to:
receive a master data set for a first spectroscopic model;
receive a target data set for a target population associated with the first spectroscopic model to update the first spectroscopic model;
generate a plurality of training data sets based on the master data set and the target data set;
generate a plurality of validation data sets based on the target data set,
wherein the plurality of validation data sets do not include data of the master data set;
determine a model setting based on the plurality of training data sets and the plurality of validation data sets and using cross-validation;
generate a second spectroscopic model based on the model setting, the target data set, and the master data set; and
provide the second spectroscopic model.

### Claim 18 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the model setting is at least one of:
a partial least squares (PLS) factor of a PLS model,
a quantity of components of a principal component regression (PCR) model,
a support vector regression (SVR) parameter of an SVR model, or
a preprocessing setting.

### Claim 19 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the one or more instructions, that cause the one or more processors to determine the model setting, cause the one or more processors to:
generate a plurality of partial performance metrics for each of the plurality of training data sets and a corresponding validation data set of the plurality of validation data sets;
aggregate the plurality of partial performance metrics to generate a total performance metric; and
determine the model setting to minimize an error value of the total performance metric.

### Claim 20 (depends on 17)

The non-transitory computer-readable medium of claim 17, wherein the one or more instructions, when executed by the one or more processors, cause the one or more processors to:
receive a measurement;
perform a spectroscopic determination based on the measurement and using the second spectroscopic model; and
provide an output identifying the spectroscopic determination.
