# US 2025/0156753 A1 · Detecting outliers during machine learning system training

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

- Number: US 2025/0156753 A1
- Document ID: US20250156753A1
- Status: Published application
- Kind: A1
- Application number: 18/543,439
- Filed: 2023-12-18
- Published: 2025-05-15
- Earliest claimed: 2023-11-15
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Lukasz G. Cmielowski; Daniel Jakub Ryszka; Szymon Kucharczyk; Dorota Laczak
- Web page: https://patentgrep.com/patent/US20250156753A1

## CPC

- G06N 20/00: Machine learning
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

One or more computer processors receiving training data comprising a trial subset of training data. The one or more computer processors probe the trial subset of training data using a machine learning system and multiple robust measures of scale formulas to select an upper bound for data outlier detection and to select a lower bound for data outlier selection. The one or more computer processors detect one or more outliers in the training data using the selected upper bound and the selected lower bound. The one or more computer processors generate modified training data using the detected outliers. The one or more computer processors train the machine learning system utilizing the modified training data.

## Claims

### Claim 1

A computer-implemented method comprising:
receiving training data comprising a trial subset of training data;
probing the trial subset of training data using a machine learning system and multiple robust measures of scale formulas to select an upper bound for data outlier detection and to select a lower bound for data outlier selection,
detecting one or more outliers in the training data using the selected upper bound and the selected lower bound;
generating modified training data using the detected outliers; and
training the machine learning system utilizing the modified training data.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein the probing comprises selecting an optimal formula from the multiple robust measures of scale formulas by:
determining a trial upper and lower bounds using at least multiple of the multiple robust measure of scale formulas;
generating trial training data using the trial upper and lower bounds for the at least multiple robust measure of scale formulas;
training the machine learning system with the trial training data for the at least multiple robust measure of scale formulas; and
determining an accuracy score for the machine learning system trained with the trial training data for the at least multiple robust measure of scale formulas using the training data; and
selecting the optimal formula from the at least multiple robust measure of scale formulas using the accuracy score.

### Claim 3 (depends on 2)

The computer-implemented method of claim 2, wherein the optimal formula comprises parameters and the probing comprises optimizing the parameters during iterative training of the machine learning system with the trial subset with variations in the parameters.

### Claim 4 (depends on 1)

The computer-implemented method of claim 1, wherein generating the modified data using the detected outliers comprises labeling the detected outliers with a missing value imputer, and the machine learning system is configured for handling the missing value imputer during training.

### Claim 5 (depends on 4)

The computer-implemented method of claim 4, wherein the missing value imputer is a not-a-number identifier.

### Claim 6 (depends on 5)

The computer-implemented method of claim 5, wherein the machine learning system is a pipeline machine learning system, and the pipeline machine learning system comprises multiple computational units arranged in a pipeline, wherein the multiple computational units comprise a transformer configured to enable or disable the effect of the missing value imputer on output of the machine learning system.

### Claim 7 (depends on 6)

The computer-implemented method of claim 6, further comprising:
generating a test group of data from the training data;
testing a first accuracy of the machine learning system with the transformer configured to enable the effect of the missing value imputer using the test group of data;
testing a second accuracy of the machine learning system with the transformer configured to disable the effect of the missing value imputer using the test group of data;
disabling the effect of the missing value imputer in the transformer if the second accuracy is greater than the first accuracy; and
enabling the effect of the missing value imputer in the transformer if the first accuracy is greater than the second accuracy.

### Claim 8 (depends on 3)

The computer-implemented method of claim 3, wherein generating modified training data using the detected outliers comprises deleting training data containing the identified outliers.

### Claim 9 (depends on 1)

The computer-implemented method of claim 1, wherein the machine learning system is an automated machine learning system, and the automated machine learning system is configured for automatically selecting an optimal machine learning module from multiple machine learning models during training of the machine learning system using the modified training data.

### Claim 10 (depends on 1)

The computer-implemented method of claim 1, wherein the multiple robust measures of scale formulas are selected from the group consisting of InterQuartile Range, Robust Covariance, Local Outlier Factor, standard deviation, median absolute deviation, median absolute deviation, Cauchy distribution, biweight midvariance.

### Claim 11

A computer program product comprising:
one or more computer readable storage media having computer-readable program instructions stored on the one or more computer readable storage media, said program instructions executes a computer-implemented method comprising steps of:
program instructions to receive training data comprising a trial subset of training data;
program instructions to probe the trial subset of training data using a machine learning system and multiple robust measures of scale formulas to select an upper bound for data outlier detection and to select a lower bound for data outlier selection,
program instructions to detect one or more outliers in the training data using the selected upper bound and the selected lower bound;
program instructions to generate modified training data using the detected outliers; and
program instructions to train the machine learning system utilizing the modified training data.

### Claim 12

A computer system comprising:
one or more computer processors;
one or more computer readable storage media having computer readable program instructions stored on the one or more computer readable storage media for execution by at least one of the one or more processors, the stored program instructions execute a computer-implemented method comprising steps of:
receiving training data comprising a trial subset of training data;
probing the trial subset of training data using a machine learning system and multiple robust measures of scale formulas to select an upper bound for data outlier detection and to select a lower bound for data outlier selection,
detecting one or more outliers in the training data using the selected upper bound and the selected lower bound;
generating modified training data using the detected outliers; and
training the machine learning system utilizing the modified training data.

### Claim 13 (depends on 12)

The computer system of claim 12, wherein the program instructions to probe comprises selecting an optimal formula from the multiple robust measures of scale formulas by:
determining a trial upper and lower bounds using at least multiple of the multiple robust measure of scale formulas;
generating trial training data using the trial upper and lower bounds for the at least multiple robust measure of scale formulas;
training the machine learning system with the trial training data for the at least multiple robust measure of scale formulas; and
determining an accuracy score for the machine learning system trained with the trial training data for the at least multiple robust measure of scale formulas using the training data; and
selecting the optimal formula from the at least multiple robust measure of scale formulas using the accuracy score.

### Claim 14 (depends on 13)

The computer system of claim 13, wherein the optimal formula comprises parameters and the probing comprises optimizing the parameters during iterative training of the machine learning system with the trial subset with variations in the parameters.

### Claim 15 (depends on 12)

The computer system of claim 12, wherein generating the modified data using the detected outliers comprises labeling the detected outliers with a missing value imputer, and the machine learning system is configured for handling the missing value imputer during training.

### Claim 16 (depends on 15)

The computer system of claim 15, wherein the missing value imputer is a not-a-number identifier.

### Claim 17 (depends on 16)

The computer system of claim 16, wherein the machine learning system is a pipeline machine learning system, and the pipeline machine learning system comprises multiple computational units arranged in a pipeline, wherein the multiple computational units comprise a transformer configured to enable or disable the effect of the missing value imputer on output of the machine learning system.

### Claim 18 (depends on 17)

The computer system of claim 17, wherein the program instructions stored on the one or more computer readable storage media, further comprise the steps of:
generating a test group of data from the training data;
testing a first accuracy of the machine learning system with the transformer configured to enable the effect of the missing value imputer using the test group of data;
testing a second accuracy of the machine learning system with the transformer configured to disable the effect of the missing value imputer using the test group of data;
disabling the effect of the missing value imputer in the transformer if the second accuracy is greater than the first accuracy; and
enabling the effect of the missing value imputer in the transformer if the first accuracy is greater than the second accuracy.

### Claim 19 (depends on 15)

The computer system of claim 15, wherein generating modified training data using the detected outliers comprises deleting training data containing the identified outliers.

### Claim 20 (depends on 12)

The computer system of claim 12, wherein the machine learning system is an automated machine learning system, and the automated machine learning system is configured for automatically selecting an optimal machine learning module from multiple machine learning models during training of the machine learning system using the modified training data.
