Army researchers have developed a new approach for training machine learning models that can better withstand dirty and deceptive data. Models trained under this method have greatly surpassed other state-of-the-art models in terms of robustness, scientists said. Machines outperform humans in many data-processing tasks, but sometimes fall victim to obvious mistakes that humans can see a mile away. Scientists at the U.S. Army Combat Capabilities Development Command's Army Research Laboratory designed a new approach that makes it harder for adversaries to trick machine learning models. "We were able to reduce model complexity by about a factor of 10 without affecting other performance metrics under benign conditions," said Army scientist Dr. Ananthram Swami.

Marketing today is on the threshold of change. In the past, marketing as we knew it was largely dominated by 30-second TV spots and other mass media such as print, outdoor, radio and so on. The number-crunching only came into play while deciding which medium to back in the advertising campaign and for what price to buy the media. But, look around today and there are the likes of Google, Facebook, Twitter and others who apply complex algorithms such as Page Rank, Adsense, marketing mix modelling, content marketing and so on along with technology (analytics, digital marketing, search engine optimisation (SEO) and search engine marketing (SEM) to make marketing a lot more data-driven. Similarly, in music the magic of maths plays a huge role.

Many believe SageMaker, the machine learning (ML) service from Amazon Web Services, has truly democratised the adoption of artificial intelligence (AI) and data science by making it available for developers, corporations, and laymen alike. What you may not know is that the idea behind this AWS offering took root in Chennai a few years ago, when a software engineer on an annual pilgrimage back home was nursing a bad bout of jet jag. Swami Sivasubramanian, VP, Machine Learning, AWS, is considered a pioneer in cloud computing. The 41-year-old joined Amazon in 2005 after completing a PhD in distributed computing, making him one of the early employees for an idea that is now a $36 billion ARR business. Over the years, he has built 40 AWS services along with his team.

The vulnerabilities of machine learning models open the door for deceit, giving malicious operators the opportunity to interfere with the calculations or decision making of machine learning systems. Scientists at the Army Research Laboratory, specializing in adversarial machine learning, are working to strengthen defenses and advance this aspect of artificial intelligence. Often, in a data set, corrupted inputs or an adversarial attack enters a machine learning model undetected. Adversaries also impact a model whether or not they know the machine learning algorithm in use, training a substitute machine learning model for use on a "victim" model. Corruption can even occur on sophisticated machine learning models trained with an abundance of data to perform critical tasks.