Singularity Pulse  
     
  Banking

AI to Help Bank Build Simulations for Making More Informed Decisions

Barclays has signed up with AI simulation firm Simudyne Technology, to help the bank make decisions across Barclays’ trading, lending and risk management divisions. Simudyne allows banks to create computer models that simulate millions of possible future scenarios, allowing them to test how individual factors will perform and interact with each other in a vast array of situations. Currently, banks largely rely on models that extrapolate data from historical data, complemented by complex and expensive simulations in niche areas such as valuing options. Justin Lyon, Simudyne chief executive says that “Given a policy change, these models can be used to see how a system will reconfigure itself — allowing machine intelligence to stay ahead of the human responses.” For lending, Simudyne’s software allows banks to build simulations based on inputs such as household incomes, how people spend their money and how they behave as borrowers.

 
  Image  
 

AI to Help Home Buyers Purchase a House

Homes.com is using AI to improve the home-search process. President David Mele stated that once the customers have picked their location, they must answer certain questions about their preferences, and then list what’s a must have, and what’s less necessary, but still nice to have. Homes.com uses a proprietary algorithm to deliver personalized, customized search results for that consumer. For instance, a buyer may need a four bedroom and three-bathroom house but want hardwood floors and a fireplace. The algorithm will return options, with a percentage of how well it matches with the customer’s needs.

Housing  
  Image  
  Defense

China Planning to Deploy Large Unmanned AI Submarines

The South China Post reported that Beijing is planning to deploy artificially intelligent unmanned submarines in the early 2020s. The seacraft could be used to survey waters, place munitions or even be used in suicide attacks against enemies. These subs, dubbed extra-large unmanned underwater vehicles(XLUUVs), are much bigger than the current crop of underwater vehicles — large enough to dock as conventional submarines and to carry significant weaponry and other equipment. Their AI will help them operate undersea, not only to avoid natural phenomena, but to detect and identify friendly or hostile ships and make navigational decisions to avoid them. The XLUUVs are also designed to complete tasks without needing to seek input during the course of a mission. One of the main advantages of the AI subs, is their relatively low cost, as all the investment that goes into making the vehicles survivable environments for humans can be stripped away.

 
  Image  
 

AI to Identify Early Stage Stomach Cancer with High Accuracy

Japanese researchers have succeeded in using artificial intelligence to identify early stage stomach cancer with a high accuracy rate. According to Riken and the National Cancer Center, it took AI only 0.004 seconds to judge whether an endoscopic image showed early stage cancer or normal stomach tissue. AI correctly detected cancer in 80 percent of cancer images, while the accuracy rate was 95 percent for normal tissue. The accuracy rates were as high as those of veteran doctors.

Healthcare  
  Image  
  Pharma

AI Robot Mixes Chemicals to Discover Reactions

Researchers have designed, built and programmed a chemical-handling robot that can screen and predict chemical reactivity using machine learning. Discovering new reactions is usually an unpredictable and time-consuming process, involving expert knowledge to target a particular molecule. The robot can perform up to 36 experiments per day – around 10 times more than a human. It builds a database of reaction information by first randomly selecting and combining different reagents from a given set. Samples of each mixture are subsequently analysed for reactivity by obtaining their spectra in real-time using built-in sensors, including NMR, infrared and mass spectrometry. A machine learning algorithm designed to recognise reactivity based on differences in spectra from the original starting reagents then classifies reaction mixtures as either reactive or unreactive. This data is then fed back to the robot to decide the next round of experiments.

 
  Image  
 

AI Transforms Data into Customer Insight for Bank

Macquarie Bank’s AI-based digital banking platform plans to create both oversight and foresight for customer accounts. This includes providing information to individual customers on their accounts and using analytics to give the bank more insight into customers. For customers, the issue of feeling that their bank does not get them or understand their needs is one of the principal reasons for dissatisfaction. Banks have failed to build up single customer view is not due to a lack of data. Instead, the problem is how to manage this data effectively at scale and across different silos.

Banking  
  Image  
  Pharma

AI Firm Developing Drugs for Genetic Disorders

Insilico Medicine and A2A Pharmaceuticals today launched Consortium.AI, a new venture founded with the goal of applying advances in artificial intelligence to cutting-edge drug discovery. Through Consortium.AI, the two companies will collaboratively develop therapeutic treatments for Duchenne Muscular Dystrophy (DMD) and other severe genetic disorders and use machine learning to validate the most promising candidates.

 
  Image  
 

Google’s Move Mirror AI Experiment Matches Poses to Photos

Move Mirror watches you move through your computer's webcam (with permission) and uses AI trickery to match your pose against a database of tens of thousands of photos. The experiment uses a machine learning model called Posenet, which recognizes the overall position of a human subject by analyzing and adding up where different parts and joints are in a photo or video. Your position is analyzed in real time and compared to a set of 80,000 photos. Move Mirror shows the closest match to each of your positions, stringing them together in a slideshow. The result is a sort of janky video of different people acting out your movements. You can even make a GIF out of it. In finding a matching image, the experiment uses the location of 17 different body parts including your shoulders, ankles and hips. It doesn't take any individual characteristics into account, such as gender, height or body type.

Miscellaneous  
  Image  
  Research

Hackers Easily Fool AIs into Seeing the Wrong Thing

AIs are surprisingly vulnerable to being spoofed. A group of researchers described a turtle they had 3D printed. Most people would say it looks just like a turtle, but an artificial intelligence (AI) algorithm saw it differently. Most of the time, the AI thought the turtle looked like a rifle. Similarly, it saw a 3D-printed baseball as an espresso. These are examples of "adversarial attacks"—subtly altered images, objects, or sounds that fool AIs without setting off human alarm bells. Computer scientists working on the attacks say they are providing a service, like hackers who point out software security flaws. According to Aleksander Madry, a computer scientist at MIT, "We need to rethink all of our machine learning pipeline to make it more robust."

 
  Image  
 
Image
“Quantum Computing, a revolution in technology, has shown great promise towards solving complex computing problems currently outside the capabilities of current computers. Though in its infancy, we at Decimal Point Analytics strongly believe that it is going to grow exponentially in the near future. It holds the potential not only to boost the AI revolution but also transform the way data is synthesized. The articles on Quantum computing will give an insight into the recent developments in this space.”
 
  Image  
  Quantum Computing

Software that Finds and Suppresses Errors in Quantum Computing

Errors are endemic in quantum computing. If error rates can be kept sufficiently low, quantum computers can offer transformational capabilities to tackle the world's most challenging computational problems. The team at Quantum Benchmark has developed software that finds and suppresses errors in the calculations performed by quantum computers and tells users the likelihood of the answer being right or wrong. Quantum Benchmark's True-Q™ is a new suite of software products, built for makers and users of quantum computers, providing industry-leading tools for error characterization, error suppression, error correction and performance validation. The True-Q software will be integrated into Google's open source platform for building quantum algorithms.

 
  Image  
 

Google's New Cirq Project Aims to Make Quantum Computers Useful

Google has launched Cirq, an open source framework for running algorithms on the quantum computers that will be available soon. A common problem researchers face when designing quantum algorithms for today’s quantum computers – the 50 to 100 qubit Noisy Intermediate-Scale Quantum devices – is in working within the limitations and nuances of the hardware. Poor mapping between the algorithms and the machines, and ignoring the devices’ complex constraints, inevitably leads to wasted resources and faulty computations. “The thesis is that when we’re developing algorithms in the NISQ era, you’re going to have to pay attention to hardware. It’s going to become very important. And the algorithm development should aim to deal with those details,” said Dave Bacon, software lead, Google AI Quantum Team.

Quantum Computing  
  Image  
 
 
 
     
  Banner  
 
Contact us:

5A, B-Wing, Trade Star Building, J. B. Nagar, Andheri-Kurla Road, Andheri (East), Mumbai - 400 059, Maharashtra, India
W: www.decimalpointanalytics.com  
E: info@decimalpointanalytics.com   T: +91 22300 15200