Inventor

CMIELOWSKI LUKASZ G

PL46 patents

⚠️ This page may combine multiple inventors who share the name “CMIELOWSKI LUKASZ G”. Patents are grouped by organization below to help distinguish them — per-person disambiguation is on the roadmap.

CMIELOWSKI LUKASZ G

IBM

Presenting discriminant change history records on topology graphs

Predicting performance of machine learning models

Metric specific machine learning model improvement through metric specific outlier removal

Machine learning model monitoring

Analyzing test result failures using artificial intelligence models

Machine learning based data monitoring

Data labeling for deep-learning models

Automatic correcting of computing cluster execution failure

Automatically determining accuracy of a predictive model

Generation of test scenarios based on risk analysis

Risk formula for erroneous software components detection

Text mining for automatically determining semantic relatedness

Analyzing test result failures using artificial intelligence models

Machine learning based data monitoring

Evaluating surrogate machine learning models

Multi-dimensional record correlations

Input feature significance identification based on batches of prediction

Dynamic recommendation system for correlated metrics and key performance indicators

Standardizing disparate data points

Deep learning experiment content generation based on single design

Continuous learning system for models without pipelines

Automatic correcting of computing cluster execution failure

Data labeling for deep-learning models

Training set creation for classifying features of a system under agile development

Prediction quality assessment

Reliable inference of a machine learning model

Regression detection and correction in relationships between performance indicators and ai metrics

Continuous learning process using concept drift monitoring

Standard error of prediction of performance in artificial intelligence model

Finding root cause for low key performance indicators

Indexing based on feature importance

Method for testing an artificial intelligence model using a substitute model

Model comparison with unknown metric importance

Early stopping of artificial intelligence model training using control limits

Multi-class classification evaluation metric determination

Exposing payload data from non-integrated machine learning systems

Assessment of machine learning performance with limited test data

Generation of test scenarios based on risk analysis

Providing a machine learning model based on desired metric values

Enhancing graph visualization with supplemental data

Enhancing graph visualization with supplemental data

Providing reduced training data for training a machine learning model

Controlling a bias of a machine learning module background

Release cycle optimization based on significant features values simulation

Reducing problem complexity when analyzing 3-D images

INT BUSINESS MACHINES CORPORATION

Testing a machine learning model