How accurate is your AI?

March 14, 2018, Kyoto University
The new AI evaluation method looks at the input data itself to find if the 'accuracy' of the AI can be trusted. Credit: Kyoto University / JB Brown

As AI's role in society continues to expand, J B Brown of the Graduate School of Medicine reports on a new evaluation method for the type of AI that predicts yes/positive/true or no/negative/false answers.

Brown's paper, published in Molecular Informatics, deconstructs the utilization of AI and analyzes the nature of the statistics used to report an AI program's ability. The new technique also generates a probability of the performance level given evaluation data, answering questions such as: What is the probability of achieving accuracy greater than 90%?

Reports of new AI applications appear in the news almost daily, including in society and science, finance, pharmaceuticals, medicine, and security.

"While reported statistics seem impressive, research teams and those evaluating the results come across two problems," explains Brown. "First, to understand if the AI achieved its results by chance, and second, to interpret applicability from the reported performance statistics."

For example, if an AI program is built to predict whether or not someone will win the lottery, it may always predict a loss. The program may achieve '99% accuracy', but interpretation is key to determine the accuracy of the conclusion that the program is accurate.

But herein lies the problem: in typical AI development, the evaluation can only be trusted if there is an equal number of positive and negative results. If the data is biased toward either value, the current system of evaluation will exaggerate the system's ability.

So to tackle this problem, Brown developed a that evaluates performance based on only the input data itself.

"The novelty of this technique is that it doesn't depend on any one type of AI technology, such as deep learning," Brown describes. "It can help develop new evaluation metrics by looking at how a metric interplays with the balance in predicted data. We can then tell if the resulting metrics could be biased."

Brown hopes this analysis will not only raise awareness of how we think about AI in the future, but also that it contributes to the development of more robust AI platforms.

In addition to the accuracy metric, Brown tested six other metrics in both theoretical and applied scenarios, finding that no single metric was universally superior. He says the key to building useful AI platforms is to take a multi-metric view of evaluation.

"AI can assist us in understanding many phenomena in the world, but for it to properly provide us direction, we must know how to ask the right questions. We must be careful not to overly focus on a single number as a measure of an AI's reliability."

Brown's program is freely available to the general public, researchers, and developers.

Explore further: Metabolic activity of brown adipose tissue easier to verify with new method

More information: J. B. Brown, Classifiers and their Metrics Quantified, Molecular Informatics (2018). DOI: 10.1002/minf.201700127

Related Stories

Choosing a simpler path to drug discovery

March 29, 2017

Drug discovery is in essence the designing of compounds to interact with disease-related proteins. And in many recent development efforts, this process increasingly relies on "big data" and complex "deep learning", requiring ...

Can 'brown fat' really help with weight loss?

November 10, 2017

There have been hyped reports of late about "brown fat". Different from the fat we mostly have (white fat), brown fat is capable of burning more energy. So the theory goes: if we have more brown fat, we can lose weight. But ...

Recommended for you

EU copyright law passes key hurdle

June 20, 2018

A highly disputed European copyright law that could force online platforms such as Google and Facebook to pay for links to news content passed a key hurdle in the European Parliament on Wednesday.

1 comment

Adjust slider to filter visible comments by rank

Display comments: newest first

rrwillsj
1 / 5 (1) Mar 14, 2018
At this point of development, I just want to hear that all these 'private enterprise', "free enterprise' 'bold innovator tycoons' are going to accept any responsibility for when things go wrong.

Or, as usual, will they have their stooge politicians sticking the liability for damages and casualties on the aching back of us sucker taxpayers?

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.