AWS ( Amazon web services ) launches new model to detect biases in machine leaning algorithms

A new feature from Amazon Web Services will alert developers to potential check the bias in machine learning algorithms, part of a larger effort by the tech industry to keep automated predictions from discriminating against women, people of colour and other underrepresented groups which was a major flaw in these machine learning algorithms.

The feature which is known as 'SageMaker Clarify' was declared at the AWS re:Invent conference on Tuesday as a new component of the AWS SageMaker machine learning platform. The technology works by analyzing the data used to train machine learning prototypes for significant signs of bias, including data sets that don’t accurately reflect the larger population. It also examines the machine learning model itself to help ensure the accuracy of the resulting predictions.

A report from the 2018 MIT study found that the presence of an irregular number of white males in data sets used to train facial identification algorithms led to a larger number of errors in recognizing women and people of a different colour, which becomes really important.

Amazon itself was reported to have discarded an artificial intelligence recruiting tool that aimed out to be biased against women, in part because the data used to train the model came from recognising patterns in past resumes submitted to the company, the majority of which were from men.

“Bias can show up at every stage of the machine learning workflow,” said Dr Nashlie Sephus, an applied science manager for Amazon Web Services AI, and this is not at all acceptable in today's scenarios, everyone should be treated equally, hence introducing the new feature at re:Invent. “So even with the best possible intentions and a whole lot of expertise, removing bias in machine learning models is difficult.” Understanding and learning machine learning becomes really important to make such complex models. Hence, it is really good to pursue your career as well in this field by pursuing machine learning certification and upskill yourself.

She quoted examples such as developing a TV show testimonial algorithm without enough television dramas in the training data. There’s also the challenge of “model drift,” where the training data end up being substantially different from the data used to make predictions, such as changes in mortgage rates causing a machine learning model for home loans to become biased.

SageMaker Clarify is one of a series of machine learning features and products announced by Swami Sivasubramanian, Amazon VP of AI, at the virtual conference Tuesday. The next major re:Invent keynote is Thursday with Peter DeSantis, vice president of global infrastructure.