Kounts advanced fraud prevention uses adaptive AI and machine learning combined with the Identity Trust Global Network to protect businesses from chargebacks and false positives

Barclaycard Payments, which processes almost 40 per cent of card transactions in the UK, has announced a new partnership with leading fraud prevention provider Kount to give Barclays Transact customers access to award-winning fraud detection software.

Barclays Transact is a suite of tools designed to help merchants make their online transactions both simpler and safer.

Transacts new fraud module, powered by Kount, uses complex data linking within the Identity Trust Global Network and artificial intelligence algorithms to detect fraudulent transactions in real-time, at the point of check-out, helping protect the business from false positives and chargebacks.

Kount is the only fraud prevention system with the Identity Trust Global Network, the largest network of trust and risk signals, which is comprised of 32 billion annual interactions from more than 6,500 customers across 75+ industries. Kounts AI utilizes both supervised and unsupervised machine learning to detect emerging and existing complex fraud. Kounts customers have reported results such as a 99 percent reduction in chargebacks, 70 percent reduction in false positives, and an 83 percent reduction in manual reviews.

Recommended AI News: Proof Of Impact Launches B2B Platform So Companies Can Track Their Impact

The fraud module can also help businesses better prepare for the introduction of the EUs Strong Customer Authentication (SCA)regulation,which aims to tackle growing rates of fraud and cybercrime. SCA requires that all EEA transactions go through a two-factor authentication process, unless they qualify for an exemption. One consequence of this regulation is that the authentication process will introduce a degree of friction into the shopper journey, which may result in an increase in cart abandonment, and ultimately in lost revenue for retailers.

The Kount and Barclaycard Payments partnership is a mutually beneficial relationship. Barclaycard Payments gains access to Kounts Identity Trust Global Network to bring insights and protection to its global merchants, especially those in Europe requiring PSD2 and SCA support. Kount broadens its global visibility with a significant increase in UK card transactions, said David Mattei, Senior Analyst at Aite Group. The real winners are merchants using Barclays Transact for advanced AI-driven fraud mitigation and prevention.

Transacts fraud module can help businesses overcome this friction by taking advantage of SCA-approved Transaction Risk Assessment (TRA) exemptions which is where transactions are judged to be sufficiently genuine, and therefore allowed to skip the two-factor authentication process up to pre-agreed thresholds.

Recommended AI News: AgUnity And Etherisc Sign MOU To Kick-Start Their Cooperation

With Kounts state-of-the-art fraud analysis, all transactions are analyzed in real time and scored on a spectrum of low to high risk. The merchants gateway then uses this score to identify the transactions which qualify for TRA exemptions. This results in a more frictionless payment journey and a faster checkout experience for customers, ultimately resulting in lower levels of basket abandonment and increased sales. Higher-risk transactions requiring further inspection will still go through two-factor authentication, or be immediately declined, in accordance with the regulation and customer risk appetite.

Brad Wiskirchen, CEO, Kount said: The eCommerce environment is rapidly growing and changing. This partnership between Barclaycard and Kount brings together both stability and innovation to provide merchants with an innovative solution to combat emerging fraud strategies while providing a seamless payment experience for customers. Barclays Transact and Kounts Identity Trust Global network operate together to deliver a top of the line customer experience. By leaving fraud prevention and regulation management up to the experts, businesses can focus on what they do best.

David Jeffrey, Director of Product, Barclaycard Payments said: We are really excited to be partnering with Kount, because they share our goal of collaborative innovation, and a drive to deliver best-in-class shopper experiences. Thanks to Kounts award-winning fraud detection software, the new module will not only help customers to fight fraud and prevent unwanted chargebacks, it will also help them to maximize sales, improve customer experience, and better prepare for the introduction of SCA.

Recommended AI News: Topgolf and Golf Scope Partner to Launch New Virtual Reality Game

See the article here:

Barclaycard Payments Partners With Kount to Deliver Industry-leading Fraud Prevention and Prepare Businesses for SCA - AiThority

A report Machine Learning has been recently published by Market Industry Reports (MIR). As per the report, the global machine learning market was estimated to be over ~US$ 2.7 billion in 2019. It is anticipated to grow at a CAGR of 43.5% from 2019 to 2030.

Major Key Players of the Machine Learning Market are: Intel, H2O.ai, Amazon Web Services, Hewlett Packard Enterprise Development LP, IBM, Google LLC, Microsoft, SAS Institute Inc., SAP SE, and BigML, Inc., among others.

Download PDF to Know the Impact of COVID-19 on Machine Learning Market at: https://www.marketindustryreports.com/pdf/133

There are various factors attributing to growth of the machine learning market including the availability of robust data sets and the adoption of machine learning techniques in modern applications such as self-driving cars, traffic alerts (Google Maps), product recommendations (Amazon), and transportation & commuting (Uber). Also, the adoption of machine learning across various industries, such as the finance industry, to minimize identity theft and detect fraud is adding to growth of the machine learning market.

Technologies powered by machine learning, capture and analyse data to improve marketing operations and enhance the customer experience. Moreover, the proliferation of large datasets, technological advancements, and techniques to provide a competitive edge in business operations are among major factors that will drive the machine learning market. Rapid urbanization, acceptance of machine learning in developed countries, rapid adoption of new technologies to minimize work and the presence of a large talent pool will push the machine learning market.

Major Applications of Machine Learning Market covered are: Healthcare & Life Sciences Manufacturing, Retail Telecommunications Government and Defense BFSI (Banking, financial services, and insurance) Energy and Utilities and Others

Research objectives:-

To study and analyze the global Machine Learning consumption (value & volume) by key regions/countries, product type and application, history data. To understand the structure of the Machine Learning market by identifying its various sub-segments. Focuses on the key global Machine Learning manufacturers, to define, describe and analyze the sales volume, value, market share, market competitive landscape, SWOT analysis, and development plans in the next few years. To analyze the Machine Learning with respect to individual growth trends, future prospects, and their contribution to the total market. To share detailed information about the key factors influencing the growth of the market (growth potential, opportunities, drivers, industry-specific challenges and risks).

Go For Interesting Discount Here:https://www.marketindustryreports.com/discount/133

Table of Content

1 Report Overview 1.1 Study Scope 1.2 Key Market Segments 1.3 Players Covered 1.4 Market Analysis by Type 1.5 Market by Application 1.6 Study Objectives 1.7 Years Considered

2 Global Growth Trends 2.1 Machine Learning Market Size 2.2 Machine Learning Growth Trends by Regions 2.3 Industry Trends

3 Market Share by Key Players 3.1 Machine Learning Market Size by Manufacturers 3.2 Machine Learning Key Players Head office and Area Served 3.3 Key Players Machine Learning Product/Solution/Service 3.4 Date of Enter into Machine Learning Market 3.5 Mergers & Acquisitions, Expansion Plans

4 Breakdown Data by Product 4.1 Global Machine Learning Sales by Product 4.2 Global Machine Learning Revenue by Product 4.3 Machine Learning Price by Product

5 Breakdown Data by End User 5.1 Overview 5.2 Global Machine Learning Breakdown Data by End User

Buy this Report @ https://www.marketindustryreports.com/checkout/133

In the end, Machine Learning industry report specifics the major regions, market scenarios with the product price, volume, supply, revenue, production, and market growth rate, demand, forecast and so on. This report also presents SWOT analysis, investment feasibility analysis, and investment return analysis.

About Market Industry Reports

Market Industry Reports is a global leader in market measurement & advisory services, Market Industry Reports is at the forefront of innovation to address the worldwide industry trends and opportunities. We identified the caliber of market dynamics & hence we excel in the areas of innovation and optimization, integrity, curiosity, customer and brand experience, and strategic business intelligence through our research.

We continue to pioneer state-of-the-art approach in research & analysis that makes complex world simpler to stay ahead of the curve. By nurturing the perception of genius and optimized market intelligence we bring proficient contingency to our clients in the evolving world of technologies, mega trends and industry convergence. We empower and inspire Vanguards to fuel and shape their business to build and grow world-class consumer products.

Contact Us- Email: [emailprotected] Phone: + 91 8956767535 Website:https://www.marketindustryreports.com

Read more:

Machine Learning Market Projected to Register 43.5% CAGR to 2030 Intel, H2Oai - Cole of Duty

Image: The Respiratory Epithelium (Photo courtesy of Wikimedia Commons)

In the study, the scientists identified ciliated cells as the major target of SARS-CoV-2 infection. The bronchial epithelium acts as a protective barrier against allergens and pathogens. Cilia removes mucus and other particles from the respiratory tract. Their findings offer insight into how the virus causes disease. The scientists infected HBECs in an air-liquid interface with SARS-CoV-2. Over a period of three days, they used single-cell RNA sequencing to identify signatures of infection dynamics such as the number of infected cells across cell types, and whether SARS-CoV-2 activated an immune response in infected cells.

The scientists utilized advanced algorithms to develop working hypotheses and used electron microscopy to learn about the structural basis of the virus and target cells. These observations provide insights about host-virus interaction to measure SARS-CoV-2 cell tropism, or the ability of the virus to infect different cell types, as identified by the algorithms. After three days, thousands of cultured cells became infected. The scientists analyzed data from the infected cells along with neighboring bystander cells. They observed ciliated cells were 83% of the infected cells. These cells were the first and primary source of infection throughout the study. The virus also targeted other epithelial cell types including basal and club cells. The goblet, neuroendocrine, tuft cells, and ionocytes were less likely to become infected.

The gene signatures revealed an innate immune response associated with a protein called Interleukin 6 (IL-6). The analysis also showed a shift in the polyadenylated viral transcripts. Lastly, the (uninfected) bystander cells also showed an immune response, likely due to signals from the infected cells. Pulling from tens of thousands of genes, the algorithms locate the genetic differences between infected and non-infected cells. In the next phase of this study, the scientists will examine the severity of SARS-CoV-2 compared to other types of coronaviruses, and conduct tests in animal models.

Machine learning allows us to generate hypotheses. Its a different way of doing science. We go in with as few hypotheses as possible. Measure everything we can measure, and the algorithms present the hypothesis to us, said senior author David van Dijk, PhD, an assistant professor of medicine in the Section of Cardiovascular Medicine and Computer Science.

Related Links:Yale School of Medicine

Follow this link:

Yale Researchers Use Single-Cell Analysis and Machine Learning to Identify Major COVID-19 Target - HospiMedica

Global Machine Learning in Medical ImagingMarket is comprehensively prepared with main focus on the competitive landscape, geographical growth, segmentation, and market dynamics, including drivers, restraints, and opportunities. This report provides a detailed and analytical look at the various companies that are working to achieve a high market share in the Global Machine Learning in Medical ImagingMarket. Data is provided for the top and fastest growing segments.

Machine Learning in Medical Imaging Market competition by top manufacturers as follow: , Zebra, Arterys, Aidoc, MaxQ AI, Google, Tencent, Alibaba,

Get a Sample PDF copy of the report @ https://reportsinsights.com/sample/13318

The global Machine Learning in Medical Imaging market has been segmented on the basis of technology, product type, application, distribution channel, end-user, and industry vertical, along with the geography, delivering valuable insights.

The Type Coverage in the Market are: , Supervised Learning, Unsupervised Learning, Reinforced Leaning

Market Segment by Applications, covers: , Breast, Lung, Neurology, Cardiovascular, Liver, Others

Market segment by Regions/Countries, this report covers North America Europe China Rest of Asia Pacific Central & South America Middle East & Africa

What does the report offer?

To get this report at a profitable rate.: https://reportsinsights.com/discount/13318

Furthermore, it offers valuable insights into the businesses for boosting the performance of the companies. Different sales and marketing approach have been mentioned to get a clear idea about how to achieve the outcomes in the industries.

The major geographical regions which include, North America, Asia Pacific, Europe, the Middle East & Africa and Latin America are studied. Top manufacturers from all these regions are studied to help give a better picture of the market investment. Production, price, capacity, revenue and many such important data is been discussed with precise data.

Most important data include the key recommendations and predictions by our analysts, intended to steer a strategic business decision. The company profiles section of this research service is a compilation of the growth strategies, financial status, product portfolio, and recent developments of key market participants. The report provides detailed industry analysis of the Global Machine Learning in Medical ImagingMarket with the help of proven research methodologies such as Porters five forces. The forces analyzed are bargaining power of the buyers, bargaining power of suppliers, threat of new entrants, threat of substitutes, and the degree of competition.

Access full Report Description, TOC, Table of Figure, Chart, etc.@ https://reportsinsights.com/industry-forecast/Machine-Learning-in-Medical-Imaging-Market-13318

About US:

Reports Insights is the leading research industry that offers contextual and data-centric research services to its customers across the globe. The firm assists its clients to strategize business policies and accomplish sustainable growth in their respective market domain. The industry provides consulting services, syndicated research reports, and customized research reports.

Contact US:

:(US) +1-214-272-0234

:(APAC) +91-7972263819

Email:info@reportsinsights.com

Sales:sales@reportsinsights.com

More:

Astonishing growth in Machine Learning in Medical Imaging Market | Competitive Analysis, Industry Dynamics, Growth Factors and Opportunities - Daily...

Machine Learning in Retail Market report is to provide accurate and strategic analysis of the Profile Projectors industry. The report closely examines each segment and its sub-segment futures before looking at the 360-degree view of the market mentioned above. Market forecasts will provide deep insight into industry parameters by accessing growth, consumption, upcoming market trends and various price fluctuations.

Machine Learning in Retail Market competition by top manufacturers as follow: , IBM, Microsoft, Amazon Web Services, Oracle, SAP, Intel, NVIDIA, Google, Sentient Technologies, Salesforce, ViSenze

Get a Sample PDF copy of the report @ https://reportsinsights.com/sample/13166

Global Machine Learning in Retail Market research reports growth rates and market value based on market dynamics, growth factors. Complete knowledge is based on the latest innovations in the industry, opportunities and trends. In addition to SWOT analysis by key suppliers, the report contains a comprehensive market analysis and major players landscape. The Type Coverage in the Market are: , Cloud Based, On-Premises

Market Segment by Applications, covers: , Online, Offline

Market segment by Regions/Countries, this report covers North America Europe China Rest of Asia Pacific Central & South America Middle East & Africa

To get this report at a profitable rate.: https://reportsinsights.com/discount/13166

Important Features of the report:

Reasons for buying this report:

Access full Report Description, TOC, Table of Figure, Chart, etc.@ https://reportsinsights.com/industry-forecast/Machine-Learning-in-Retail-Market-13166 About US:

Reports Insights is the leading research industry that offers contextual and data-centric research services to its customers across the globe. The firm assists its clients to strategize business policies and accomplish sustainable growth in their respective market domain. The industry provides consulting services, syndicated research reports, and customized research reports.

Contact US:

:(US) +1-214-272-0234

:(APAC) +91-7972263819

Email:info@reportsinsights.com

Sales:sales@reportsinsights.com

More:

Covid-19 Positive Impact on Machine Learning in Retail Market 2020-2025 Country Level Analysis, Current Trade Size And Future Prospective - Daily...

Last week, OpenAI published a paper detailing GPT-3, a machine learning model that achieves strong results on a number of natural language benchmarks. At 175 billion parameters, where a parameter affects datas prominence in an overall prediction, its the largest of its kind. And with a memory size exceeding 350GB, its one of the priciest, costing an estimated $12 million to train.

A system with over 350GB of memory and $12 million in compute credits isnt hard to swing for OpenAI, a well-capitalized company that teamed up with Microsoft to develop an AI supercomputer. But its potentially beyond the reach of AI startups like Agolo, which in some cases lack the capital required. Fortunately for them, experts believe that while GPT-3 and similarly large systems are impressive with respect to their performance, they dont move the ball forward on the research side of the equation. Rather, theyre prestige projects that simply demonstrate the scalability of existing techniques.

I think the best analogy is with some oil-rich country being able to build a very tall skyscraper, Guy Van den Broeck, an assistant professor of computer science at UCLA, told VentureBeat via email. Sure, a lot of money and engineering effort goes into building these things. And you do get the state of the art in building tall buildings. But there is no scientific advancement per se. Nobody worries about the U.S. is losing its competitiveness in building large buildings because someone else is willing to throw more money at the problem. Im sure academics and other companies will be happy to use these large language models in downstream tasks, but I dont think they fundamentally change progress in AI.

Indeed, Denny Britz, a former resident on the Google Brain team, believes companies and institutions without the compute to match OpenAI, DeepMind, and other well-funded labs are well-suited to other, potentially more important research tasks like investigating correlations between model sizes and precision. In fact, he argues that these labs lack of resources might be a good thing because it forces them to think deeply about why something works and come up with alternative techniques.

There will be some research that only [tech giants can do], but just like in physics [where] not everyone has their own particle accelerator, there is still plenty of other interesting work, Britz said. I dont think it necessarily creates any imbalance. It doesnt take opportunities away from the small labs. It just adds a different research angle that wouldnt have happened otherwise. Limitations spur creativity.

OpenAI is a counterpoint. It has long asserted that immense computational horsepower in conjunction with reinforcement learning is a necessary step on the road to AGI, or AI that can learn any task a human can. But luminaries like Milafounder Yoshua Bengio and Facebook VP and chief AI scientist Yann LeCunargue that AGI is impossible to create, which is why theyre advocating for techniques like self-supervised learning and neurobiology-inspired approaches that leverage high-level semantic language variables. Theres also evidence that efficiency improvements might offset the mounting compute requirements; OpenAIs own surveys suggestthat since 2012, the amount of compute needed to train an AI model to the same performance on classifying images in a popular benchmark (ImageNet) has been decreasing by a factor of two every 16 months.

The GPT-3 paper, too, hints at the limitations of merely throwing more compute at problems in AI. While GPT-3 completes tasks from generating sentences to translating between languages with ease, it fails to perform much better than chance on a test adversarial natural language inference that tasks it with discovering relationships between sentences. A more fundamental [shortcoming] of the general approach described in this paper scaling up any model is that it may eventually run into (or could already be running into) the limits of the [technique], the authors concede.

State-of-the-art (SOTA) results in various subfields are becoming increasingly compute-intensive, which is not great for researchers who are not working for one of the big labs, Britz continued. SOTA-chasing is bad practice because there are too many confounding variables, SOTA usually doesnt mean anything, and the goal of science should be to accumulate knowledge as opposed to results in specific toy benchmarks. There have been some initiatives to improve things, but looking for SOTA is a quick and easy way to review and evaluate papers. Things like these are embedded in culture and take time to change.

That isnt to suggest pioneering new techniques is easy. A 2019 meta-analysis of information retrieval algorithms used in search engines concluded the high-water mark was actually set in 2009. Another study in 2019 reproduced seven neural network recommendation systems and found that six failed to outperform much simpler, non-AI algorithms developed years before, even when the earlier techniques were fine-tuned. Yet another paper found evidence that dozens of loss functions the parts of algorithms that mathematically specify their objective had not improved in terms of accuracy since 2006. And a study presented in March at the 2020 Machine Learning and Systems conference found that over 80 pruning algorithms in the academic literature showed no evidence of performance improvements over a 10-year period.

But Mike Cook, an AI researcher and game designer at Queen Mary University of London, points out that discovering new solutions is only a part of the scientific process. Its also about sussing out where in society research might fit, which small labs might be better able determine because theyre unencumbered by the obligations to which privately backed labs, corporations, and governments are beholden. We dont know if large models and computation will always be needed to achieve state-of-the-art results in AI, Cook said. [In any case, we] should be trying to ensure our research is cheap, efficient, and easily distributed. We are responsible for who we empower, even if were just making fun music or text generators.

See more here:

OpenAIs massive GPT-3 model is impressive, but size isnt everything - VentureBeat

IEDs and so-called butterfly landminescould be detected over wide areas using drones and advanced machine learning, according to research from Binghamton University, State University at New York.

The team had previously developed a method that allowed for the accurate detection of butterfly landmines using low-cost commercial drones equipped with infrared cameras.

EPSRC-funded project takes dual approach to clearing landmines

Their new research focuses on automated detection of landmines using convolutional neural networks (CNN), which they say is the standard machine learning method for object detection and classification in the field of remote sensing. This method is a game-changer in the field, said Alek Nikulin, assistant professor of energy geophysics at Binghamton University.

All our previous efforts relied on human-eye scanning of the dataset, Nikulin said in a statement.Rapid drone-assisted mapping and automated detection of scatterable mine fields would assist in addressing the deadly legacy of widespread use of small scatterable landmines in recent armed conflicts and allow to develop a functional framework to effectively address their possible future use.

There are at least 100 million military munitions and explosives of concern devices in the world, of various size, shape and composition. Furthermore,an estimated twenty landmines are placed for every landmine removed in conflict regions

Millions of these are surface plastic landmines with low-pressure triggers, such as the mass-produced Soviet PFM-1 butterfly landmine. Nicknamed for their small size and butterfly-like shape, these mines are extremely difficult to locate and clear due to their small size, low trigger mass and a design that mostly excluded metal components, making them virtually invisible to metal detectors.

The design of the mine combined with a low triggering weight have earned it notoriety as the toy mine, due to a high casualty rate among small children who find these devices while playing and who are the primary victims of the PFM-1 in post-conflict nations, like Afghanistan.

The researchers believe that these detection and mapping techniques are generalisable and transferable to other munitions and explosives. They could be adapted to detect and map disturbed soil for improvised explosive devices (IEDs).

The use of Convolutional Neural Network-based approaches to automate the detection and mapping of landmines is important for several reasons, the researchers said in a paper published inRemote Sensing. One, it is much faster than manually counting landmines from an orthoimage (i.e. an aerial image that has been geometrically corrected). Two, it is quantitative and reproducible, unlike subjective human error-prone ocular detection. And three, CNN-based methods are easily generalisable to detect and map any objects with distinct sizes and shapes from any remotely sensed raster images.

More here:

Butterfly landmines mapped by drones and machine learning - The Engineer

Computer scientists working for a high-tech division of Google are testing how machine learning algorithms can be created from scratch, then evolve naturally, based on simple math.

Experts behind Google's AutoML suite of artificial intelligence tools have now showcased fresh research which suggests the existing software could potentially be updated to "automatically discover" completely unknown algorithms while also reducing human bias during the data input process.

Read more

According to ScienceMag, the software, known as AutoML-Zero, resembles the process of evolution, with code improving every generation with little human interaction.

Machine learning tools are "trained" to find patterns in vast amounts of data while automating such processes and constantly being refined based on past experience.

But researchers say this comes with drawbacks that AutoML-Zero aims to fix. Namely, the introduction of bias.

"Human-designed components bias the search results in favor of human-designed algorithms, possibly reducing the innovation potential of AutoML," their team's paper states. "Innovation is also limited by having fewer options: you cannot discover what you cannot search for."

The analysis, which was published last month on arXiv, is titled "Evolving Machine Learning Algorithms From Scratch" and is credited to a team working for Google Brain division.

"The nice thing about this kind of AI is that it can be left to its own devices without any pre-defined parameters, and is able to plug away 24/7 working on developing new algorithms," Ray Walsh, a computer expert and digital researcher at ProPrivacy, told Newsweek.

As noted by ScienceMag, AutoML-Zero is designed to create a population of 100 "candidate algorithms" by combining basic random math, then testing the results on simple tasks such as image differentiation. The best performing algorithms then "evolve" by randomly changing their code.

The resultswhich will be variants of the most successful algorithmsthen get added to the general population, as older and less successful algorithms get left behind, and the process continues to repeat. The network grows significantly, in turn giving the system more natural algorithms to work with.

Haran Jackson, the chief technology officer (CTO) at Techspert, who has a PhD in Computing from the University of Cambridge, told Newsweek that AutoML tools are typically used to "identify and extract" the most useful features from datasetsand this approach is a welcome development.

"As exciting as AutoML is, it is restricted to finding top-performing algorithms out of the, admittedly large, assortment of algorithms that we already know of," he said.

"There is a sense amongst many members of the community that the most impressive feats of artificial intelligence will only be achieved with the invention of new algorithms that are fundamentally different to those that we as a species have so far devised.

"This is what makes the aforementioned paper so interesting. It presents a method by which we can automatically construct and test completely novel machine learning algorithms."

Jackson, too, said the approach taken was similar to the facts of evolution first proposed by Charles Darwin, noting how the Google team was able to induce "mutations" into the set of algorithms.

"The mutated algorithms that did a better job of solving real-world problems were kept alive, with the poorly-performing ones being discarded," he elaborated.

"This was done repeatedly, until a set of high-performing algorithms was found. One intriguing aspect of the study is that this process 'rediscovered' some of the neural network algorithms that we already know and use. It's extremely exciting to see if it can turn up any algorithms that we haven't even thought of yet, the impact of which to our daily lives may be enormous." Google has been contacted for comment.

The development of AutoML was previously praised by Alphabet's CEO Sundar Pichai, who said it had been used to improve an algorithm that could detect the spread of breast cancer to adjacent lymph nodes. "It's inspiring to see how AI is starting to bear fruit," he wrote in a 2018 blog post.

The Google Brain team members who collaborated on the paper said the concepts in the most recent research were a solid starting point, but stressed that the project is far from over.

"Starting from empty component functions and using only basic mathematical operations, we evolved linear regressors, neural networks, gradient descent... multiplicative interactions. These results are promising, but there is still much work to be done," the scientists' preprint paper noted.

Walsh told Newsweek: "The developers of AutoML-Zero believe they have produced a system that has the ability to output algorithms human developers may never have thought of.

"According to the developers, due to its lack of human intervention AutoML-Zero has the potential to produce algorithms that are more free from human biases. This theoretically could result in cutting-edge algorithms that businesses could rely on to improve their efficiency.

"However, it is worth bearing in mind that for the time being the AI is still proof of concept and it will be some time before it is able to output the complex kinds of algorithms currently in use. On the other hand, the research [demonstrates how] the future of AI may be algorithms produced by other machines."

View post:

Artificial Intelligence That Can Evolve on Its Own Is Being Tested by Google Scientists - Newsweek

Silicon Integration Initiativehas launched an industry-wide survey to identify planned usage and structural gaps for prioritizing and implementing artificial intelligence and machine learning in semiconductor electronic design automation.

The Si2 platform provides a unique opportunity for semiconductor companies, EDA suppliers and IP providers to voice their needs and focus resources on common solutions, including enabling and leveraging university research.

The survey is organized by a recently formed Si2 Special Interest Group chaired by Joydip Das, senior engineer, Samsung Electronics, and co-chaired by Kerim Kalafala, senior technical staff member, EDA, and master inventor, IBM. The 18-member group will identify where industry collaboration will help eliminate deficiencies caused by a lack of common languages, data models, labels, and access to robust and categorized training data.

Recommended AI News:Artio Medical Appoints Jeff Weinrich to Board of Directors

This SIG is open to all Si2 members. Current members include:

Advanced Micro Devices ANSYS Cadence Design Systems Hewlett Packard Enterprise IBM Intel Intento Design Keysight Technologies Mentor, a Siemens Business

NC State University PFD Solutions

Qualcomm

Samsung Electronics

Sandia National Laboratories Silvaco

Synopsys Thrace Systems Texas Instruments

The survey is open April 15 May 15.

Leigh Anne Clevenger, Si2 senior data scientist, said that the survey results would help prioritize SIG activities and timelines. The SIG will identify and develop requirements for standards that ensure data and software interoperability, enabling the most efficient design flows for production, Clevenger said. The ultimate goal is to remove duplicative work and the need for data model translators, and focus on opening avenues for breakthroughs from suppliers and users alike.

Recommended AI News:Ligandal Is Developing Potential Antidote And Vaccine To SARS-CoV-2

High manufacturing costs and the growing complexity of chip development are spurring disruptive technologies such as AI and ML, Clevenger explained. The Si2 platform provides a unique opportunity for semiconductor companies, EDA suppliers and IP providers to voice their needs and focus resources on common solutions, including enabling and leveraging university research.

See the article here:

si2 Launches Survey on Artificial Intelligence and Machine Learning in Eda - AiThority

Machine learning has fundamentally changed how we engage with technology. Today, it's able to curate social media feeds, recognize complex images, drive cars down the interstate, and even diagnose medical conditions, to name a few tasks.

But while machine learning technology can do some things automatically, it still requires a lot of input from human engineers to set it up, and point it in the right direction. Inevitably, that means human biases and limitations are baked into the technology.

So, what if scientists could minimize their influence on the process by creating a system that generates its own machine-learning algorithms? Could it discover new solutions that humans never considered?

To answer these questions, a team of computer scientists at Google developed a project called AutoML-Zero, which is described in a preprint paper published on arXiv.

"Human-designed components bias the search results in favor of human-designed algorithms, possibly reducing the innovation potential of AutoML," the paper states. "Innovation is also limited by having fewer options: you cannot discover what you cannot search for."

Automatic machine learning (AutoML) is a fast-growing area of deep learning. In simple terms, AutoML seeks to automate the end-to-end process of applying machine learning to real-world problems. Unlike other machine-learning techniques, AutoML requires relatively little human effort, which means companies might soon be able to utilize it without having to hire a team of data scientists.

AutoML-Zero is unique because it uses simple mathematical concepts to generate algorithms "from scratch," as the paper states. Then, it selects the best ones, and mutates them through a process that's similar to Darwinian evolution.

AutoML-Zero first randomly generates 100 candidate algorithms, each of which then performs a task, like recognizing an image. The performance of these algorithms is compared to hand-designed algorithms. AutoML-Zero then selects the top-performing algorithm to be the "parent."

"This parent is then copied and mutated to produce a child algorithm that is added to the population, while the oldest algorithm in the population is removed," the paper states.

The system can create thousands of populations at once, which are mutated through random procedures. Over enough cycles, these self-generated algorithms get better at performing tasks.

"The nice thing about this kind of AI is that it can be left to its own devices without any pre-defined parameters, and is able to plug away 24/7 working on developing new algorithms," Ray Walsh, a computer expert and digital researcher at ProPrivacy, told Newsweek.

If computer scientists can scale up this kind of automated machine-learning to complete more complex tasks, it could usher in a new era of machine learning where systems are designed by machines instead of humans. This would likely make it much cheaper to reap the benefits of deep learning, while also leading to novel solutions to real-world problems.

Still, the recent paper was a small-scale proof of concept, and the researchers note that much more research is needed.

"Starting from empty component functions and using only basic mathematical operations, we evolved linear regressors, neural networks, gradient descent... multiplicative interactions. These results are promising, but there is still much work to be done," the scientists' preprint paper noted.

Related Articles Around the Web

View post:

New AI improves itself through Darwinian-style evolution - Big Think