Singularity Pulse  
     
   
 

University Associated – Sun Yat-sen University

University Associated – Sun Yat-sen University
In order to evaluate sentiments on Chinese financial markets, the researchers At Cornell University have crawled news and comments from multiple influential websites and analyzed the same using techniques of NLP. The “Senti-scores” were calculated using techniques like tokenization, Word2vec word embedding and semantic database WordNet. Researchers made use of selenium to crawl the news. After the initial downloading of news, it was subjected to pre-treatment before passing it through the algorithms to calculate sentimental scores. Within sentimental factors, researches built a finance-specific sentimental lexicon in order to ensure that sentimental factors are reflecting sentiments of financial markets specifically, and not general sentiments like happiness or sadness. 50,000 common words from Chinese vocabulary and 3,000+ words from news articles were analyzed to get Senti-score of words in advance. The algorithm computed sentimental factors of the Chinese market and the correlation coefficient was found to be a significant one. The model scored an accuracy of 73% on 450+ words which had determining sentimental effect. The algorithms used for this research are available on Github as an open source code for reference.

Read More...
 
  Image  
  Research

AI Strategy Mimics How Brains Learn to Smell

Scientists are studying the sense of smell or olfaction to gain and improve their understanding of how organisms process chemical information, code them to translate into a proper response. ML techniques built on visuals as input picks out small, well-defined features like edges, textures, colors, involving spatial mapping. Vision and olfaction use different strategies to deal with different types of data. Smells are unstructured. Unlike objects, they can’t be grouped in space. Odors are analyzed by a shallow, three-layer network that’s considerably less complex than the visual cortex. Olfaction works better when it comes to speed of learning. An olfactory-based search algorithm performed two to three times better than traditional nonbiological methods.

 
  Image  
 

Robots to Use AI in Combat Modules

Russian arms manufacturer Kalashnikov Concern has claimed to develop an AI algorithm that can differentiate between a friend and an enemy in a warzone. The robot makes use of its AI technology to hold or fire on the target accordingly. It is completely autonomous. The combat module scans an area and identifies threats, indicates the type of object (person or machine), decides on the necessary number of shots for guaranteed destruction and then carries out an attack. It is equipped with a gyro-stabilized camera system to be able to shoot while on the move. The AI makes it able to not only recognize targets, but also select priority targets. The sophisticated optics allows it to operate both in daytime and in absence of it.

Defense  
  Image  
  Research

AI to Measure Ad Content Reaction of Consumers

University Associated: Imperial College London
Realeyes, an emotion AI firm, have developed AI technology to gauge amount of attention their advertisements will receive from their target audience even before the launch of a particular advertisement. The algorithm uses webcams to capture consumer behavior traits like eye movements, blinking, yawning and head movements. It then creates metrics showing the volume and quality of attention to the ad content shown to them. The company has collaborated with AI experts from Imperial College London to develop this solution.

 
  Image  
 

Algorithm to Predict Genetic Traits by Analysing DNA

Scientists from Michigan State University have built a program which is capable of analyzing a patient’s complete genome structure. It is also capable of accurately predicting their height with an error of margin of about 2.5 centimeters. The machine-learning system was trained on a dataset of nearly 500,000 adults. It also predicted results on bone density and ultimate level of education attained. The accuracy level of height predictions has been the best when compared to other two prediction categories. Researchers state that the tool could be used to predict health risks or to create personalized drug therapies. It could also be used to screen embryos and create "designer babies."

HealthCare  
  Image  
  HealthCare

Early Detection of Alzheimer’s by AI

Researchers from the University of Toronto, Canada have designed an algorithm that learns signatures from magnetic resonance imaging (MRI), genetics and clinical data to make accurate prediction whether a person’s cognitive decline will lead to Alzheimer’s disease in the next five years. The algorithm was trained using data from more than 800 people. The data set included people ranging from normal healthy seniors to those experiencing mild cognitive impairment, and Alzheimer’s disease patients. Researchers are hopeful that the algorithm will enable them to detect signs even farther into the future.

 
  Image  
 

AI Detects the Smallest of Earthquake Activity

Earthquakes that score 2.0 or less magnitude on the moment magnitude scale are difficult to detect as a result of background noise, small events or wrongly treated as false positives. Department of Geophysics at Stanford University have developed an AI system titled, “Cnn-Rnn Earthquake Detector” (CRED) which can identify seismic signals from historical and continuous data. The algorithm consists of neural network layers of two types: convolutional neural networks and recurrent neural networks. The former extracts feature from seismographs, while the latter combines memory and inputs to make predictions. Data recorded in Arkansas, during 2011 containing 3,788 events, along with 889 monitoring stations was used to train the program. The learned model achieved “superior” performance compared to two widely deployed seismic systems

Miscellaneous  
  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

Researchers Creates ‘Quantum Artificial Life’ for the First Time

A team of researchers have made use of a quantum computer to create artificial life for the first time. It is a simulation of living organisms which will enable scientists to study them at the cellular level. In a quantum computer, living organisms were represented by superconducting qubits and were made to “mate”, interact with environment, and “die”. This activity was used to develop a model of major factors that influences evolution. The aim of the modeling exercise is to create a computer model that replicates the processes of Darwinian evolution on a quantum computer. They used five qubit quantum processor.

 
  Image  
 

First Time Public Access to a Quantum Computer

D Wave Systems has opened up access to its systems to the public with launch of its new Leap Quantum Application Environment (QAE). It provides users with real time access over cloud to D-Wave’s quantum computers. The free access will last for an entire minute within which developers can submit and run the applications on the system and see the results in just seconds. In a span of mere sixty seconds, it can solve between 400 to 4000 problems. The range of tools available includes source development tools, interactive demos and coding examples, educational resources and knowledge base articles for users to make use of.

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