# US 10,309,894 B2 · Identification using spectroscopy

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

- Number: US 10,309,894 B2
- Document ID: US10309894B2
- Status: Granted
- Kind: B2
- Application number: 15/247,554
- Filed: 2016-08-25
- Published: 2017-03-02
- Granted: 2019-06-04
- Earliest claimed: 2015-08-26
- Assignees: Viavi Solutions Inc.
- Applicants: Viavi Solutions Inc.
- Inventors: Changmeng Hsiung; Christopher G. Pederson; Peng Zou; Lan Sun
- Web page: https://patentgrep.com/patent/US10309894B2

## CPC

- 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)}
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/00: Machine learning
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- G16C 20/20: Identification of molecular entities, parts thereof or of chemical compositions
- G16C 20/90: Programming languages; Computing architectures; Database systems; Data warehousing
- G16C 20/70: Machine learning, data mining or chemometrics
- 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)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G16C: COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE

## Abstract

A device may receive information identifying results of a spectroscopic measurement of an unknown sample. The device may perform a first classification of the unknown sample based on the results of the spectroscopic measurement and a global classification model. The device may generate a local classification model based on the first classification. The device may perform a second classification of the unknown sample based on the results of the spectroscopic measurement and the local classification model. The device may provide information identifying a class associated with the unknown sample based on performing the second classification.

## Claims

### Claim 1

A device, comprising:
a memory; and
one or more processors to:
receive a global classification model from another device,
the global classification model being generated by the other device and being distributed to a plurality of devices,
the plurality of devices including the device, and
the global classification model including a set of classes;
receive information identifying results of a spectroscopic measurement of an unknown sample;
perform a first classification of the unknown sample based on the results of the spectroscopic measurement and the global classification model;
generate a local classification model based on the first classification,
the local classification model including a subset of the set of classes;
perform a second classification of the unknown sample based on the results of the spectroscopic measurement and the local classification model; and
provide information identifying a class associated with the unknown sample based on performing the second classification.

### Claim 2 (depends on 1)

The device of claim 1,
where the one or more processors are further to:
determine a set of respective probabilities associated with the set of classes,
a particular probability, of the set of respective probabilities, indicating a likelihood that the unknown sample is associated with a particular class of the set of classes, and
select the subset of the set of classes based on the set of respective probabilities; and
where the one or more processors, when generating the local classification model, are to:
generate the local classification model based on the subset of the set of classes.

### Claim 3 (depends on 1)

The device of claim 1, where the one or more processors are further to:
perform an autoscaling pretreatment procedure; and
perform at least one of the first classification or the second classification based on performing the autoscaling pretreatment procedure.

### Claim 4 (depends on 1)

The device of claim 1,
where the other device is associated with a first spectrometer;
where the global classification model is generated using one or more spectroscopic measurements performed by the first spectrometer;
where the one or more processors are further to:
cause the spectroscopic measurement to be performed by a second spectrometer,
the second spectrometer being different from the first spectrometer; and
where the one or more processors, when performing the first classification of the unknown sample, are to:
perform the first classification of the unknown sample based on the global classification model generated using the one or more spectroscopic measurements performed by the first spectrometer and based on the results of the spectroscopic measurement performed by the second spectrometer.

### Claim 5 (depends on 1)

The device of claim 1,
where a set of classes of the global classification model correspond to a set of compounds and the class is included in the set of classes; and
where the one or more processors, when providing information identifying the class, are to:
provide information identifying a compound, of the set of compounds, corresponding to the class.

### Claim 6 (depends on 1)

The device of claim 1,
where the one or more processors, when performing the second classification, are to:
determine that a spectrum associated with the unknown sample is associated with the class,
the spectrum being identified by the results of performing the spectroscopic measurement; and
where the one or more processors, when providing information identifying the class, are to:
provide information identifying the class based on determining that the spectrum associated with the unknown sample is associated with the class.

### Claim 7 (depends on 1)

The device of claim 1,
where a support vector machine (SVM) classifier technique is utilized to generate at least one of the global classification model or the local classification model; and
where the SVM classifier technique is associated with at least one of:
a radial basis function type of kernel function,
a linear function type of kernel function,
a sigmoid function type of kernel function,
a polynomial function type of kernel function, or
an exponential function type of kernel function.

### Claim 8 (depends on 1)

The device of claim 1, where the one or more processors, when performing the second classification, are to:
assign the unknown sample to the class based on at least one of:
a probability value, or
a decision value.

### Claim 9

A 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 global classification model from another device,
the global classification model being distributed by the other device to a plurality of devices,
the plurality of devices including the device, and
the global classification model including a set of classes;
receive information identifying results of a set of spectroscopic measurements of an unknown set,
the unknown set including a plurality of unknown samples;
perform a first classification of the plurality of unknown samples based on the results of the set of spectroscopic measurements and the global classification model,
the global classification model utilizing a support vector machine (SVM) linear classifier technique;
generate a set of local classification models for the plurality of unknown samples based on the first classification,
the set of local classification models utilizing the SVM linear classifier technique, and
the set of local classification models including a subset of classes of the set of classes of the global classification model;
perform a second classification of the plurality of unknown samples based on the results of the set of spectroscopic measurements and the set of local classification models; and
provide information identifying classifications of the plurality of unknown samples based on performing the second classification.

### Claim 10 (depends on 9)

The computer-readable medium of claim 9,
where the global classification model is generated based on one or more spectroscopic measurements performed by a first spectrometer;
where the one or more instructions, that cause the one or more processors to receive the information identifying the results of the set of spectroscopic measurements, cause the one or more processors to:
receive the information identifying the results of the set of spectroscopic measurements from a second spectrometer,
the second spectrometer being different from the first spectrometer; and
where the one or more instructions, that cause the one or more processors to perform the first classification, cause the one or more processors to:
perform the first classification using the results of the set of spectroscopic measurements received from the second spectrometer and the global classification model generated based on the one or more spectroscopic measurements performed by the first spectrometer.

### Claim 11 (depends on 9)

The computer-readable medium of claim 9, where the one or more instructions, that cause the one or more processors to receive the information identifying the results of the set of spectroscopic measurements, cause the one or more processors to:
receive a plurality of spectra corresponding to the plurality of unknown samples; and
where the one or more instructions, that cause the one or more processors to perform the first classification, cause the one or more processors to:
assign the plurality of spectra to one or more classes of the global classification model,
the one or more classes of the global classification model corresponding to one or more compounds.

### Claim 12 (depends on 11)

The computer-readable medium of claim 11,
where the one or more instructions, when executed by the one or more processors, further cause the one or more processors to:
determine one or more confidence metrics for the one or more classes and a particular spectrum of the plurality of spectra,
a confidence metric, of the one or more confidence metrics, indicating a likelihood that the particular spectrum is associated with a particular class, of the one or more classes, corresponding to the confidence metric; and
assign the particular spectrum to the particular class based on the one or more confidence metrics; and
where the one or more instructions, that cause the one or more processors to generate the set of local classification models, cause the one or more processors to:
select a subset of the one or more classes based on the one or more confidence metrics; and
generate a particular local classification model, of the set of local classification models, based on the subset of the one or more classes.

### Claim 13 (depends on 9)

The computer-readable medium of claim 9,
where the one or more instructions, that cause the one or more processors to receive the information identifying the results of the set of spectroscopic measurements, cause the one or more processors to:
receive a plurality of spectra corresponding to the plurality of unknown samples;
where the one or more instructions, that cause the one or more processors to perform the second classification, cause the one or more processors to:
assign the plurality of spectra to one or more classes of the set of local classification models,
the one or more classes of the set of local classification models corresponding to one or more compounds; and
where the one or more instructions, that cause the one or more processors to provide information identifying the classifications of the plurality of unknown samples, are further to:
provide information indicating a class, of the one or more classes, to which a spectrum, of the plurality of spectra, associated with an unknown sample, of the plurality of unknown samples, is assigned based on assigning the plurality of spectra to the one or more classes.

### Claim 14 (depends on 9)

The computer-readable medium of claim 9, where the one or more instructions, that cause the one or more processors to perform the second classification, cause the one or more processors to:
determine a set of decision values associated with a particular local classification model, of the set of local classification models, and a particular unknown sample, of the plurality of unknown samples,
a decision value corresponding to a class of a set of classes of the particular local classification model; and
assign the particular unknown sample to the class of the set of classes based on the set of decision values.

### Claim 15

A method, comprising:
receiving, by a device, a global classification model from another device,
the global classification model being generated by the other device and being distributed to a plurality of devices,
the plurality of devices including the device, and
the global classification model including a set of classes;
receiving, by the device, information identifying results of a spectroscopic measurement of an unknown sample performed by a first spectrometer;
performing, by the device, a first classification of the unknown sample based on the results of the spectroscopic measurement and the global classification model,
the global classification model being generated utilizing a support vector machine (SVM) classifier technique and a set of spectroscopic measurements performed by a second spectrometer associated with the other device;
generating, by the device, a local classification model based on the first classification,
the local classification model utilizing the SVM classifier technique, and
the local classification model including a subset of classes of the set of classes of the global classification model;
performing, by the device, a second classification of the unknown sample based on the results of the spectroscopic measurement and the local classification model; and
providing, by the device, information identifying a class, of the subset of classes, associated with the unknown sample based on performing the second classification.

### Claim 16 (depends on 15)

The method of claim 15, where a kernel function associated with the SVM classifier technique includes at least one of:
a radial basis function type of kernel function,
a linear function type of kernel function,
a sigmoid function type of kernel function,
a polynomial function type of kernel function, or
an exponential function type of kernel function.

### Claim 17 (depends on 15)

The method of claim 15, where performing the second classification comprises:
assigning the unknown sample to the class, of the subset of classes, based on at least one of:
a probability value associated with the class, or
a decision value associated with the class.

### Claim 18 (depends on 15)

The method of claim 15, where the first spectrometer is different from the second spectrometer.

### Claim 19 (depends on 15)

The method of claim 15, further comprising:
storing the global classification model via a data structure,
where performing the first classification comprises:
obtaining the global classification model from the data structure; and
performing the first classification using the global classification model.

### Claim 20 (depends on 15)

The method of claim 15, further comprising:
providing information identifying a confidence metric associated with the second classification,
the confidence metric representing a measure of confidence with which the unknown sample is assigned to the class.