After Facebook announced its own tool to detect bias in an algorithm earlier this month, a new report suggests that Microsoft is also building a tool to automate the identification of bias in a range of different Artificial Intelligence (AI) algorithms.
The Microsoft tool has the potential to help businesses make use of AI without inadvertently discriminating against certain groups of people, MIT Technology Review reported on Friday.
Although Microsoft’s new tool may not eliminate the problem of bias that may creep into Machine-Learning models altogether, it will help AI researchers catch more instances of unfairness, Rich Caruna, a senior researcher at Microsoft who is working on the bias-detection dashboard, was quoted as saying.
“Of course, we can’t expect perfection — there’s always going to be some bias undetected or that can’t be eliminated — the goal is to do as well as we can,” he said.
The issue of bias will become crucial as more customers make use of these algorithms to take important decisions.
At its annual developer conference on May 2, Facebook announced its own bias-catching tool, called Fairness Flow, as the social network has found that the number of people using AI to make important decisions is increasing at the company. (IANS)
With the incidence of ovarian cancer likely to increase by 55 per cent in another 15 years or so, researchers have created an artificial intelligence (AI) software to help best treat ovarian cancer that will pave the way for personalised medicine and expedite relief, a new study says.
The mathematical software tool — TEXLab — can also predict what treatment might be most effective for patients with the World Ovarian Cancer Coalition predicting that deaths will likely increase by 67 per cent by 2035 due to this particular cancer.
The technology can be used to identify patients who are unlikely to respond to standard treatments and offer alternatives as ovarian cancer is the sixth most common cancer in women in the UK that usually strikes after menopause or those with a family history of the disease.
Early detection of the disease could improve survival rates, the study noted.
“Long-term survival rate for patients with advanced ovarian cancer is poor despite advancements in treatments. There is an urgent need for new ways,” said lead author Eric Aboagye, Professor at Imperial College London.
For the study, researchers used the software to identify the aggressiveness of tumours in CT scans and tissue samples from 364 women with ovarian cancer.
The patients were then given a score known as Radiomic Prognostic Vector (RPV) which indicates how severe the disease is, ranging from mild to severe.
The findings, published in Nature Communications, showed that the software was up to four times more accurate for predicting deaths from ovarian cancer than standard methods.
In addition, five per cent of patients with high RPV scores had a survival rate of less than two years, results showed.
High RPV was also associated with chemotherapy resistance and poor surgical outcomes, suggesting that RPV can be used as a potential bio-marker to predict how patients would respond to treatments.
“Our technology is able to give clinicians more detailed and accurate information on how the patients are likely to respond to different treatments, which could enable them to make better and more targeted treatment decisions,” said Aboagye.