Singularity Pulse  
     
  Banking

AI to Reduce Bank Fraud Detection Time to Less Than Two Seconds

As mobile banking grows, so does monetary fraud base on one's mobile phone. Microsoft plans to implement an AI-based model which can reduce bank fraud detection times to less than two seconds. The AI solution Microsoft has built and deployed on the cloud promises to identify and stop fraudulent transactions under two seconds. How does the AI detect a fraud transaction? By processing incoming and outgoing mobile phone activity and building a behavioral profile for every single bank's customer who installs the bank's app -- and flagging irregular activity in the space of less than two seconds.

 
  Image  
 

AI Technology to Teach Robots to Manipulate Objects

OpenAI, San Francisco-based startup, has developed a model that teaches robotic hand to manipulate objects. Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have discussed a computer vision system that allows robots to inspect, visually understand, and then manipulate object they’ve never seen before. It is based on dubbed dense object nets. Existing algorithms are unable to make robotic hands grasp objects if they are in a different orientation. The newly developed algorithm is a self-supervised deep neural network layered algorithm that mimics the functions of neurons in the brain.

Information Technology  
  Image  
  Banking

Neural Network-Based Models to Predict Defaults

ANZ bank is currently developing methodology to better predict defaults amongst its customers. The methodology is based on neural network-based models. It is being developed in partnership with Nvidia and Monash University. Usually, three to six months were required to develop a new risk model, but with neural network based system, the same is being developed within weeks. Currently, the bank is testing its model for biases in its decision making algorithms.

 
  Image  
 

Google Releases Free Māori Translation App

A mobile app called, “Kupu” deploys photo recognition technology to identify items in “te reo”, also known as Māori language. Dean Mahuta, Māori language expert, served as an advisor for the project. Google’s AI technologies were used on Te Aka Māori Dictionary, an online source for data with 300000 plus visitors per month. The algorithm is designed to learn and update itself based on the feedback it receives from users for each translation done.

Information Technology  
  Image  
  Research

Extracting Keywords from Open-Ended Business Survey Questions

University Associated: Cornell University
The cost of analyzing free-text survey data is more than collecting and analyzing survey data with Multiple Choice questions (MCQs). Yet, the analysis is important as contains new content beyond predefined categories. Researchers at Cornell University have published a paper wherein these human interpretations and opinions are analyzed using Natural Language Processing (NLP) technology. It automates the analysis of natural language and provides with insights on human judgements. The code developed is kept open-source and is generalizable to other datasets.

 
  Image  
 

Banking Sector Set to Improve Its Operational Efficiency

Arya.ai is one of the front runners in Indian markets which is aimed to reduced processing time for banking, financial services and insurance (BFSI) sector. Founded by IIT Mumbai alumni, the company’s core product, “Vega” a ‘Deep Learning platform’ in banking offers a complete environment to test and train many use cases without the hassle of having to reinvent the stack repeatedly. And with optimized training, the total time from designing a network to deploying the solution is reduced further. The product is used for cheque automation, scanning images of cheques and their clearance. It is also currently helping hospitals clear their insurance claims faster. Traditionally, parameters like room rent, claim reports, policy details, etc., are considered for clearance, costing a lot of time. The same can be done in minutes via company’s AI program.

Banking  
  Image  
  Miscellaneous

ML Algorithm to Classify Cooking Dishes and Recipes

AI technology deep convolutional neural networks (CNN) has been used by a GitHub user to classify images into food categories and also to give a matching recipe as an output. The dataset for training the deep-CNN was taken from chefkoch.de, the largest German-language dataset. The dataset contains more than 400000 food images and 300000+ recipes for analyzing. Object recognition using CNN and next-neighbor classification techniques were used over those images.

 
  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

World’s First Quantum Computing App for Protein Folding

Rigetti Computing, leading full-stack quantum computing platforms has gone into a partner program named, “ProteinQure” to develop first generation of practical quantum applications. Under this partnership, publicly available tools will be developed to conduct further research in field of quantum computing and biology. Currently, company has folded the largest proteins on two different quantum computing platforms namely; quantum annealers as well as universal quantum computers. It is hopeful of building world’s first quantum app for protein folding using software library “Forest”.

 
  Image  
 

A Quantum Computing
Based ETF

Defiance ETFs have announced launch of “Defiance Quantum ETF (QTUM)”, a fund which will focus on providing investors with access to technologies focused on increasing computer performance speed and efficiency. With the launch of this fund, investors need not wait aside to gain exposure to promising technology of future. QTUM will offer clients to invest in a transparent way in companies involved in developing ML and Quantum Computing technologies. Few of the sectors covered will be hardware companies, software makers and defense and security firms.

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