Posts Tagged ‘neural networks’

Plumbing The Depths Of Deep Learning

Image Courtesy: Flickr/ Hey Paul Studios

Image Courtesy: Flickr/ Hey Paul Studios

Will deep learning take us where we want to go? It’s one of the questions that Oxford University professor of Computational Linguistics Stephen Pulman will be delving into at this week’s Sentiment Analysis Symposium. There, he’ll be participating in a workshop session today on compositional sentiment analysis and giving a presentation tomorrow on bleeding-edge natural language processing.

“There is a lot of hype about deep learning, but it’s not a magic solution,” says Pulman. “I worry whenever there is hype about some technologies like this that it raises expectations to the point where people are bound to be disappointed.”

That’s not to imply, however, that important progress isn’t taking place when it comes to deep learning, which leverages machine learning methods based on learning representations with applications to everything from NLP to computer vision to speech recognition.

Read more

Meet Spaun: The Future of Artificial Intelligence

spaun

Naomi Eterman of McGill Daily recently discussed a technology developed in 2012 by scientists at the University of Waterloo: “Spaun, short for Semantic Pointer Architecture Unified Network, is the largest computer simulation of a functioning brain to date. It is the brainchild of Chris Eliasmith, a professor in philosophy and systems design engineering at the University of Waterloo, who developed the system as a proof-of-principle supplement to his recent book: How to Build a Brain. The model is composed of 2.5 million simulated neurons and four different neurotransmitters that allow it to ‘think’ using the same kind of neural connections as the mammalian brain. Read more

Bing Turns to Deep Learning to Improve Image Search

honey

Zach Walton of Web Pro News recently wrote, “Image search is a cornerstone of any search engine. That’s why both Google and Bing are doing everything they can to improve image search to bring up the most relevant images for any search imaginable. While some may argue that recent changes made to Google image search make it worse, Bing is moving ahead with a new strategy that involves deep learning. So, what is deep learning? In short, it’s a type of machine learning that uses artificial neural networks to learn about and understand multiple concepts, including the abstract. In the past, computer systems had to be manually ‘trained’ to recognize patterns or specific images. With machine learning, these systems can now learn to recognize these patterns on their own. When it comes to image search quality, Bing found that integrating deep learning into its systems greatly increased the quality.” Read more

Gatfol Announces Successful Closing of Angel Seed Fund Round, EU Collaboration and Reaching a Hundred Thousand Users per Month

gatfol

Luqa, Malta (PRWEB UK) 21 November 2013 — As part of its development initiative into Europe, Gatfol today announced the reaching of further milestones: On reaching set technological programming benchmarks, Gatfol received final funding installments in its initial seed fund round with current investors. Gatfol will utilise the funds to set full EU development- and marketing functionality with Malta as base and operational satellites in New Zealand for Asia-Pacific reach and through government funded programs in South Africa for African continent expansion. Read more

Delfigo Security Awarded US Patent on Event-Driven AI Architecture

BOSTON, Sept. 12, 2013 /PRNewswire-iReach/ — Delfigo Security filed for United States Patent protection (US Serial No: 12/221,757) on August 6, 2008 for Combining Artificial Intelligence (AI) Concepts with an Event-Driven Security Architecture. Since that date, Delfigo has developed and implemented key concepts described in this original patent, such as cloud and mobile authentication as well as intelligent mobile fraud detection.  This key patent was awarded to Delfigo on September 12, 2013. Read more

EverString Maps Hidden Relationships Between Internet Companies

Megan Ross Dickey of SF Gate recently wrote, “How do all the companies on the Internet fit together? Where is the center? What is on the outside and moving in? Menlo Park, Ca. startup EverString has put the whole thing on a map, using artificial intelligence technology to perform semantic analysis on 900 million news sentences and counting. The company was started last September by former investment banker Vincent Yang, who are both students at the Stanford Graduate School of Business. Their team includes a neural network brain scientist, NASA scientists, mathematicians, growth equity investors, event trading experts, and a professional gambler, many of them also MBA students at Stanford.” Read more

Determining Context on Mobile Devices

A new article on EDN Network by Debbie Meduna, Dev Rajnarayan, and Jim Steele of Sensor Platforms, Inc. discusses how to make mobile platforms context-aware. The team writes, “Using the sensors available on these platforms, one can infer the context of the device, its user, and its environment.  This can enable new useful applications that make smart devices smarter.  It is common to utilize machine learning to determine the underlying meaning in the large amount of sensor data being generated all around us. However, traditional machine learning methods (such as neural networks, polynomial regression, and support vector machines) do not directly lend themselves to application in a power-conscious mobile environment.  The necessary techniques for ensuring effective implementation of sensor context are discussed.” Read more

Nara Offers Restaurant Recommendations Anywhere in the U.S.

CAMBRIDGE, Mass.–(BUSINESS WIRE)–June 04, 2013– Nara Logics Inc., a company that solves the problem of Web search by crafting a more personalized and liberating Web, today announced the availability of its service nationwide. Previously available in 50 cities throughout the United States and Canada, Nara is now able to curate a unique set of personalized restaurant recommendations for the individual anywhere across the country, from big cities to small towns and everywhere in between. Additionally, Nara has launched its Native iOS app for a sleek, user-friendly mobile experience, as well as a new Website redesign. Read more

Throwing Some Semantic Fun Into the April Fool’s Web Mix

Image Courtesy Flickr/ Sean MacEntee

It’s April Fool’s Day on the Web, and we’re sensing some semantic allusions and downright sentiment analytics assertions in today’s pranks. Have a look:

  • Head over to your Google search engine and you’ll be teased to find out what that smell is with Google Nose. or, as they describe it, the new scentsation in search.  Go beyond type, talk, and touch for a new notation of sensation, it promises. The Internet sommelier, Google explains, comes with an expertly curated Knowledge Panels to pair images, descriptions, and aromas. While it credits new technologies such as StreetSense (responsible for Google inhaling and indexing millions of atmospheric miles), and Android Ambient Odor Detection (which collects smells via the mobile OS), it seems to me that the Knowledge Graph had to have a hand in this one.

Semantic Tech Turns Up Biomarkers And Phenotypes, Avoids Dead Ends And Higher Costs

Image Courtesy: ipharmd.net

Dr. Carlo Trugenberger, co-founder and Chief Scientific Officer at InfoCodex Semantic Technologies AG, has co-authored a report reflecting the topic he discussed at last fall’s London SemTech event: An approach to drug research that relies on identifying relevant biochemical information using the company’s autonomous self-organizing semantic engines to text mine large repositories of biomedical research papers.

The model, says Trugenberger, is a departure from many other semantically-engineered approaches to streamlining drug research, which are based on natural language processing (NLP). That’s good for extracting information from documents, he says, but not as adept at discovering knowledge. “That’s what our InfoCodex software is designed for, to find new facts and hidden correlations” in repositories of unstructured information.

Read more

NEXT PAGE >>