Archives for algorithms have flaws

16 Jun

IBM Launches FreaAI That Automatically Finds Weaknesses In ML models

image-23601
image-23601

IBM’s Research team at Haifa built FreaAI to detect flaws in machine learning models by automatically scrutinising human-interpretable chunks of data to predict when algorithms are working and when they are not.  The researchers addressed the challenge of finding under-performing data slices to validate the ML performance of ML solutions. They developed automated slicing heuristics…

The post IBM Launches FreaAI That Automatically Finds Weaknesses In ML models appeared first on Analytics India Magazine.

16 Jun

IBM Launches FreaAI That Automatically Finds Weaknesses In ML models

image-23602
image-23602

IBM’s Research team at Haifa built FreaAI to detect flaws in machine learning models by automatically scrutinising human-interpretable chunks of data to predict when algorithms are working and when they are not.  The researchers addressed the challenge of finding under-performing data slices to validate the ML performance of ML solutions. They developed automated slicing heuristics…

The post IBM Launches FreaAI That Automatically Finds Weaknesses In ML models appeared first on Analytics India Magazine.