Archive for the ‘Machine Learning’ Category

Adversarial attacks in machine learning: What they are and how to stop them – VentureBeat

Elevate your enterprise data technology and strategy at Transform 2021.

Adversarial machine learning, a technique that attempts to fool models with deceptive data, is a growing threat in the AI and machine learning research community. The most common reason is to cause a malfunction in a machine learning model. An adversarial attack might entail presenting a model with inaccurate or misrepresentative data as its training, or introducing maliciously designed data to deceive an already trained model.

As the U.S. National Security Commission on Artificial Intelligences 2019 interim report notes, a very small percentage of current AI research goes toward defending AI systems against adversarial efforts. Some systems already used in production could be vulnerable to attack. For example, by placing a few small stickers on the ground, researchers showed that they could cause a self-driving car to move into the opposite lane of traffic. Other studies have shown that making imperceptible changes to an image can trick a medical analysis system into classifying a benign mole as malignant, and that pieces of tape can deceive a computer vision system into wrongly classifying a stop signas a speed limit sign.

The increasing adoption of AI is likely to correlate with a rise in adversarial attacks. Its a never-ending arms race, but fortunately, effective approaches exist today to mitigate the worst of the attacks.

Attacks against AI models are often categorized along three primary axes influence on the classifier, the security violation, and their specificity and can be further subcategorized as white box or black box. In white box attacks, the attacker has access to the models parameters, while in black box attacks, the attacker has no access to these parameters.

An attack can influence the classifier i.e., the model by disrupting the model as it makes predictions, while a security violation involves supplying malicious data that gets classified as legitimate. A targeted attack attempts to allow a specific intrusion or disruption, or alternatively to create general mayhem.

Evasion attacks are the most prevalent type of attack, where data are modified to evade detection or to be classified as legitimate. Evasion doesnt involve influence over the data used to train a model, but it is comparable to the way spammers and hackers obfuscate the content of spam emails and malware. An example of evasion is image-based spam in which spam content is embedded within an attached image to evade analysis by anti-spam models. Another example is spoofing attacks against AI-powered biometric verification systems..

Poisoning, another attack type, is adversarial contamination of data. Machine learning systems are often retrained using data collected while theyre in operation, and an attacker can poison this data by injecting malicious samples that subsequently disrupt the retraining process. An adversary might input data during the training phase thats falsely labeled as harmless when its actually malicious. For example, large language models like OpenAIs GPT-3 can reveal sensitive, private information when fed certain words and phrases, research has shown.

Meanwhile, model stealing, also called model extraction, involves an adversary probing a black box machine learning system in order to either reconstruct the model or extract the data that it was trained on. This can cause issues when either the training data or the model itself is sensitive and confidential. For example, model stealing could be used to extract a proprietary stock-trading model, which the adversary could then use for their own financial gain.

Plenty of examples of adversarial attacks have been documented to date. One showed its possible to 3D-print a toy turtle with a texture that causes Googles object detection AI to classify it as a rifle, regardless of the angle from which the turtle is photographed. In another attack, a machine-tweaked image of a dog was shown to look like a cat to both computers and humans. So-called adversarial patterns on glasses or clothing have been designed to deceive facial recognition systems and license plate readers. And researchers have created adversarial audio inputs to disguise commands to intelligent assistants in benign-sounding audio.

In apaper published in April, researchers from Google and the University of California at Berkeley demonstrated that even the best forensic classifiers AI systems trained to distinguish between real and synthetic content are susceptible to adversarial attacks. Its a troubling, if not necessarily new, development for organizations attempting to productize fake media detectors, particularly considering the meteoric riseindeepfakecontent online.

One of the most infamous recent examples is Microsofts Tay, a Twitter chatbot programmed to learn to participate in conversation through interactions with other users. While Microsofts intention was that Tay would engage in casual and playful conversation, internet trolls noticed the system had insufficient filters and began feeding Tay profane and offensive tweets. The more these users engaged, the more offensive Tays tweets became, forcing Microsoft to shut the bot down just 16 hours after its launch.

As VentureBeat contributor Ben Dickson notes, recent years have seen a surge in the amount of research on adversarial attacks. In 2014, there were zero papers on adversarial machine learning submitted to the preprint server Arxiv.org, while in 2020, around 1,100 papers on adversarial examples and attacks were. Adversarial attacks and defense methods have also become a highlight of prominent conferences including NeurIPS, ICLR, DEF CON, Black Hat, and Usenix.

With the rise in interest in adversarial attacks and techniques to combat them, startups like Resistant AI are coming to the fore with products that ostensibly harden algorithms against adversaries. Beyond these new commercial solutions, emerging research holds promise for enterprises looking to invest in defenses against adversarial attacks.

One way to test machine learning models for robustness is with whats called a trojan attack, which involves modifying a model to respond to input triggers that cause it to infer an incorrect response. In an attempt to make these tests more repeatable and scalable, researchers at Johns Hopkins University developed a framework dubbed TrojAI, a set of tools that generate triggered data sets and associated models with trojans. They say that itll enable researchers to understand the effects of various data set configurations on the generated trojaned models and help to comprehensively test new trojan detection methods to harden models.

The Johns Hopkins team is far from the only one tackling the challenge of adversarial attacks in machine learning. In February, Google researchers released apaper describing a framework that either detects attacks or pressures the attackers to produce images that resemble the target class of images. Baidu, Microsoft, IBM, and Salesforce offer toolboxes Advbox, Counterfit, Adversarial Robustness Toolbox, and Robustness Gym for generating adversarial examples that can fool models in frameworks like MxNet, Keras, Facebooks PyTorch and Caffe2, Googles TensorFlow, and Baidus PaddlePaddle. And MITs Computer Science and Artificial Intelligence Laboratory recently released a tool called TextFoolerthat generates adversarial text to strengthen natural language models.

More recently, Microsoft, the nonprofit Mitre Corporation, and 11 organizations including IBM, Nvidia, Airbus, and Bosch releasedtheAdversarial ML Threat Matrix, an industry-focused open framework designed to help security analysts to detect, respond to, and remediate threats against machine learning systems. Microsoft says it worked with Mitre to build a schema that organizes the approaches malicious actors employ in subverting machine learning models, bolstering monitoring strategies around organizations mission-critical systems.

The future might bring outside-the-box approaches, including several inspired by neuroscience. For example, researchers at MIT and MIT-IBM Watson AI Lab have found that directly mapping the features of the mammalian visual cortex onto deep neural networks creates AI systems that are more robust to adversarial attacks. While adversarial AI is likely to become a never-ending arms race, these sorts of solutions instill hope that attackers wont always have the upper hand and that biological intelligence still has a lot of untapped potential.

Read more from the original source:
Adversarial attacks in machine learning: What they are and how to stop them - VentureBeat

Relogix Announces Collaboration with Dr. Graham Wills, Predictive Analytics and Machine Learning Expert, To Better Predict Office Space Needs -…

Relogix will be the first in the industry to more accurately forecast and predict companies' real estate needs. Companies will potentially save hundreds of millions of real estate spend, year over year with this collaborative innovation between Relogix and Dr. Wills. "Relogix has a significant data set to work with, from years of collecting billions of terabytes of Corporate Real Estate data around the world," says Dr. Wills. "I'm excited to use this data and cutting-edge machine learning techniques to take spatial data research to the next level."

With the pandemic, it has become ever more difficult for companies to understand workplace demand for real estate, with everyone working from home and anywhere for the foreseeable future. As people return to the office, understanding the relationship between people and their demand for workspace is a significant challenge for workplace technology leaders in Corporate Real Estate, HR, and IT.

"We're making a significant R&D investment to further innovation around forecasting and predictive analytics for Corporate Real Estate," says Andrew Millar, Founder and CEO of Relogix. "We are excited to be working with Graham, a pre-eminent researcher in the AI field, and expect our collaboration to leverage advanced machine learning techniques to surface insights like never before."

As an outstanding data science leader for over 20 years, Wills is a disruptive innovator, who has been innovating predictive analytics and forecasting for 30 years. Hailing from IBM, Dr. Wills is a well-known researcher in the fields of spatial data exploration and time series monitoring. At IBM, Wills was the lead architect for predictive analytics and machine learning in IBM's Data and AI group, and led the development of major advances including intelligent automatic forecasting, natural language data insights, anomaly detection and key driver identification.

About Graham Wills, PhD:Graham's passion is analyzing data and designing capabilities that help others do the same with their data. His focus is on creating software systems that allow non-experts to draw conclusions safely and efficiently from predictive and machine learning models, and thus enhance the value of their data. Graham has authored over 60 publications, including a book in the Springer statistical series, and has chaired or presented at numerous international statistical and knowledge discovery conferences. His patents span visualization, spatial analysis, semantic knowledge, and associated AI domains. Graham believes that the goal of AI is to give professionals the assistance they need to make great decisions from their data, and that CRE is an ideal domain in which to introduce new AI and Machine Learning capabilities to revolutionize the marketplace.

About Andrew Millar, CEO:Andrew's mission is to turn data into valuable outcomes. With over 20 years as a corporate real estate solutions and insights provider, Relogix founder and CRE veteran, Andrew Millar, recognized the need for technology in the CRE industry. He founded Relogix out of a need to create solutions to help organizations evolve their workspace and get high quality data to drive strategic decision making. Andrew believes that the key to evolving workspace and strategic planning lies in data science. Just like the workplace, data science is progressive: it is a journey of perpetual discovery, refinement, and adaptation. Andrew has since created proprietary sensor technology with the needs of corporate real estate in mind technology created for CRE professionals by CRE professionals.

About Relogix:Trusted by top Corporate Real Estate professionals who need to make data-driven business decisions to inform their real estate strategy and measure impact. Our flexible workplace insights platform and state-of-the-art IoT occupancy sensors are proven to transform the workplace experience. We're always looking for the next innovation in workplace technology, leveraging two decades of CRE and analytics expertise to help our clients understand and optimize their global real estate portfolios.

SOURCE Relogix Inc.

Read the rest here:
Relogix Announces Collaboration with Dr. Graham Wills, Predictive Analytics and Machine Learning Expert, To Better Predict Office Space Needs -...

Man, meet machine: the role of AI and machine learning in the modern sales desk | Global Banking & Finance Review – Global Banking And Finance…

By Matthew Hodgson, CEO, Mosaic Smart Data

In our last article we looked at how productivity is one of the core benefits of a bank gaining control over its data and analysing it more effectively. But once a bank has gained this control and insight, how does it go a step further by augmenting sales teams with AI and machine learning tools that can digest large data sets and alert them to the needs of their clients?

Traditionally, banks have always held an information advantage over their clients in fact their business models have been specifically designed to leverage the market information to which they have access and transform it into value enhancing insight.

But in recent years, this advantage has been slowly chipped away at as the markets have become increasingly electronic and the buy-side has upped its game in terms of the data to which it has access, and its ability to analyse large amounts of it. As market and price transparency has increased, one of the core competitive advantages of a banks sales desk has been eroded.

Against this shifting backdrop, banks are starting to realise the potential power of innovative machine learning and AI tools in helping them to upskill and maintain their competitiveness in the sales arena. There is a dawning realisation that backward looking BI analysis is not fit for purpose in driving business forward, especially in this era of utilising AI to squeeze every possible efficiency and productivity from the resources at hand.

Some are now starting to deploy these technologies to enable predictive and prescriptive analytics, as well as connecting systems to prompt them as to the next best action for their clients. By absorbing information that might otherwise be missed, AI delivers the analysis to drive new sales engagement with clients and by delivering those insights at the optimum time.

Investment and adoption at scale is expected to increase significantly over the coming years. This comes as no surprise when you consider it has been estimated by McKinsey that AI can potentially unlock $1 trillion of incremental value for banks[1]. These tools can be thought of as a GPS for the sales desk those banks without it will struggle to compete against more forward-leaning firms who are empowering their employees with the most advanced digital tools.

The evolving role of the salesperson

According to a recent report from PwC, almost 80% of banking and capital markets CEOs see skills shortages as a threat to their growth prospects.[2]This is because, quite simply, banks havent managed to keep pace with the changing manner in which their clients want to interact with them.

While no one is suggesting robots will completely replace salespeople any time in the near future, there are certain skills that can be enhanced when man and machine work together in tandem. One of the main skills that clients increasingly demand from banks is a more customised and tailored experience, which in turn drives a more intimate and refined relationship.

In addition, as electronification continues to grow, sales teams tend to manage a larger pool of clients across asset classes. Clients expect salespeople to provide a seamless service in multiple asset classes and have a global view of flows across the organisation.

Data therefore needs to be aggregated from across the organisation and made available to salespeople in one consolidated and comprehensive view so they can, for example, alert clients about new investment opportunities as they unfold no matter the asset class.

Bridging the skills gap

In recent years, a growing number of large investment banks have launched ambitious projects to apply AI and machine learning techniques to previously unexplored data sources, in order to bridge the skills gap and improve how they sell to clients. A recent survey found that 75% of banks with over $100 billion in assets are currently implementing AI strategies.[3]

Using the right technology, a combination of internal transaction data, external data feeds and unstructured data sources such as newsfeeds, can be standardised and aggregated into one holistic view. AI-powered advisory tools can then be applied to help banks anticipate client activity in order to build inventory for expected demand, identify unique and unforeseen market opportunities, extract timely information from news and websites, and alert sales based on market triggers.

Using AI and machine learning you can, for example, see which customers are likely to defect and move their business elsewhere, and therefore up your defensive measures. After all, it is much more expensive to acquire a new customer than it is to maintain an existing one. You can also become more responsive and relevant to clients, because you are able to see what customer activity you anticipate on a particular day and then serve that customer with the appropriate inventory.

This technology has been leveraged over the last number of years to improve the service high-street banks deliver to retail customers. However, within investments banks the benefits of these same tools are beneficial to sales desks covering all types of clients including corporates, hedge funds, asset managers, insurers, pension funds, central banks and even internal clients.

Some banks are also exploring the use of natural language generation (NLG). This is a software process that automatically transforms data into a written narrative, making lightning-fast generation of expert business intelligence and reporting a reality in todays financial markets.

NLG can generate intuitive prose that reads as if it were written by the best quant in the house at the click of a button, equipping sales teams with the collateral they need to offer up the most appropriate trading opportunities to their clients. These reports can even be prepared with enough variance and nuance in language and style to keep the copy fresh and engaging to the reader. This power of NLG is driving enormous time saving benefits across the organisation by taking laborious daily tasks and automating them at the click of a button.

Becoming AI-first

These are just a handful of examples of how AI and machine learning can help sales desks deepen customer relationships, provide personalised insights and recommendations, and, ultimately, turn the profit dial in their favour.

Banks that fail to make AI central to their core strategy and operationsoften referred to as becoming AI-firstwill risk being overtaken by competition and deserted by their customers in the coming years.

The current operating environment is both uncertain and challenging for investment banks, but a carefully planned programme that builds on cutting-edge data analytics and AI technology holds the key to driving growth and delivering the modern, information-driven trading experience that clients demand.

After all, its typically during periods of stress where relationships are forged. As a bank, if youre able to guide a client through the fog of confusion, you will likely have a relationship for life and AI and machine learning can assist in facilitating this.

But dont just take our word for it. A client recently told us that since deploying AI technology across the front desk, their sales team had made 20% more calls, had 22% longer conversations with clients, and this had resulted in significantly more volume seen and executed. If youre a salesperson known to have the best information, the client will call you first. Its that simple.

Visit link:
Man, meet machine: the role of AI and machine learning in the modern sales desk | Global Banking & Finance Review - Global Banking And Finance...

Company uses AWS, genomics and machine learning to develop a blood test for early cancer detection – TechRepublic

Hospitals and businesses use cloud computing, machine learning and voice-controlled devices to personalize healthcare for patients.

Image: 3dreams/Shutterstock

Personalizing healthcare requires the power of cloud computing whether the challenge is screening for cancer, reducing the paperwork load for doctors or making decisions about care, according to speakers at the AWS Healthcare and Life Sciences Virtual Symposium.

Wilson To, the worldwide head of healthcare at AWS, hosted the event at the end of May. To and four guests discussed how cloud services can improve information management to personalize healthcare.

Josh Ofman, chief medical officer for Grail, said that his company is using cloud computing to detect cancer at earlier stages when it is easier to treat. The Galleri test uses a blood test to screen for multiple cancers at once.

Ofman said that genomics and machine learning are the foundation of the new early detection test. The test looks for epigenetic changes in a person's DNA that can be a warning sign for mutations caused by cancer.

According to the company, the test has a false positive rate of less than 0.5% and a positive predictive value of 44%.

Grail recommends the Galleri test for people 50 and older who are at a higher risk of cancer. The company also suggests that the test be used in addition to other screenings, not as a replacement for existing procedures. The company claims that the test can identify more than 50 types of cancer ranging from Hodgkin and non-Hodgkin lymphoma, melanoma and soft tissue sarcoma.

Grail started working with AWS in 2017 to ingest and analyze hundreds of thousands of records and genomic datasets. Grail migrated its core processing and analytical infrastructure from on-premises to a cloud platform at that time. Grail uses storage, compute and network services from AWS.

"This collaboration is powering our growth and will enable us to get to scale," Ofman said.

SEE: Cloud data storage policy (TechRepublic Premium)

Ofman said the company's data set will grow by orders of magnitude as researchers process all the samples they have today.

"It will enable us to continue to refine our test and develop new products in new disease areas," he said.

According to the National Cancer Institute, the most common cancers in men are prostate, lung and colorectal cancers, which make up about 43% of cancers diagnosed in men in 2020. For women, the types that represented 50% of all cancer diagnoses in 2020 were breast, lung and colorectal.

The retail cost of the test is $949. According to the company, the test is not covered by insurance.

Three other AWS customers spoke at the event, including Biogen, Cambia Health Solutions and Houston Methodist Hospital. Laurent Rotival, chief information officer and senior vice president at Cambia Health Solutions, said his company uses AWS to bring together data streams from disparate sources to create a coherent experience for customers.

Alisha Alaimo, president of Biogen's U.S. organization, explained how the company worked with Us Against Alzheimer's to develop a screening test. The idea was to make the test feel more personalized and less intimidating.

The brain health test can be taken by an individual with concerns for herself, or by a caregiver who is worried about a loved one. The screening is at Mybrainguide.org and is anonymous and available in English and Spanish.

Roberta Schwartz, chief innovation officer and executive vice president of Houston Methodist Hospital, described the health system's work with Alexa and voice commands to improve patient care. Schwartz also sees a need for more personalized healthcare services, a trend that the pandemic intensified. The hospital system used these guidelines to revamp the patient experience: Help me now, make it easy and remember me.

Another goal of the project was to let doctors have more face time than screen time when working with patients.

The hospital has Amazon Echos in every room and Schwartz said she has seen a new level of acceptance of the devices among patients and doctors.

"The devices were essential when patients couldn't have visitors," she said. "We are planning to hook our Alexas up to the nurse call system as well."

The hospital also plans to use the devices to reduce the time doctors have to spend transcribing patient information and to make it easier to pull up relevant information during a patient consultation.

During a 34-week pilot program, the hospital deployed 1,200 devices in its facilities and saw more than 600 daily interactions with Alexa and Avia, a virtual health assistant. Requests for music were the most popular request at 75% followed by knowledge searches, socializing, inquiries about the weather and general communication.

This is your go-to resource for XaaS, AWS, Microsoft Azure, Google Cloud Platform, cloud engineering jobs, and cloud security news and tips. Delivered Mondays

Visit link:
Company uses AWS, genomics and machine learning to develop a blood test for early cancer detection - TechRepublic

Machine Learning Market 2021 Global Industry Forecasts Analysis, Competitive Landscape and Key Regions Analysis The Manomet Current – The Manomet…

The latest Global Machine Learning Marketreport lends a competitive head start to businesses by offering accurate predictions for this vertical at both regional and global scale. It entails a top-to-bottom evaluation of the various industry segments, highlighting the current and future development possibilities, and all other factors affecting the revenue potential. Moreover, the research piece covers the leading companies, as well emerging contenders and newcomers to provide a holistic view of the competitive landscape. Additionally, it makes inclusion of the challenges due to the Covid-19 pandemic and the potential paths going forward.

Ask for a Sample PDF Copy of Machine Learning Market with Figures, Graphs and Tocs:https://www.bigmarketresearch.com/request-sample/3024215?utm_source=MMT&utm_medium=MWA

The process begins with internal and external sources to obtain qualitative and quantitative information related to the Machine Learning Market. It also provides an overview and forecast for the Machine Learning Market based on all the segmentation provided for the global region. The predictions highlighted in the Machine Learning Market share report have been derived using verified research procedures and assumptions. By doing so, theBig Market Researchreport serves as a repository of analysis and information for every component of the Machine Learning Market

NOTE:Our report highlights the major issues and hazards that companies might come across due to the unprecedented outbreak of COVID-19.

Market players have been discussed and profiles of leading players including Top Key Companies:International Business Machines CorporationMicrosoft CorporationSAP SESas Institute Inc.Amazon Web Services, Inc.Bigml, Inc.Google Inc.Fair Isaac CorporationBaidu, Inc.Hewlett Packard Enterprise Development LpIntel CorporationH2o.ai

Talk to our Analyst / Ask for a discount on Machine Learning Market and Get More Information Related to This Report:https://www.bigmarketresearch.com/request-for-discount/3024215?utm_source=MMT&utm_medium=MWA

Key highlights from Covid-19 impact assessment:

Why to Select This Report:

The Machine Learning Market is also characterized by a highly complex value chain involving product manufacturers, material suppliers, technology developers, and manufacturing equipment developers. Partnerships between research organizations and the industry players help in streamlining the path from the lab to commercialization. In order to also leverage the first mover benefit, companies need to collaborate with each other so as to develop products and technologies that are unique, innovative and cost effective.

How will this Market Intelligence Report Benefit You?

The report includes the region-wise segmentation North America (United States, Canada and Mexico), Europe (Germany, France, UK, Russia and Italy), Asia-Pacific (China, Japan, Korea, India and Southeast Asia), South America (Brazil, Argentina, Colombia etc.), Middle East and Africa (Saudi Arabia, UAE, Egypt, Nigeria and South Africa) of the market. In the regional segmentation, the regions dominating the Machine Learning market are included along with the regions where the growth of the market is slow.

By the product type, the Machine Learning Market is primarily split into 2020-2025:CloudOn-premises

By the end-users/application, the Machine Learning Market report covers the following segments 2020-2025:BFSIHealthcare and Life SciencesTelecommunicationManufacturingOthers

Conclusively, this report is a one stop reference point for the industrial stakeholders to get Machine Learning market forecast of till 2025. This report helps to know the estimated market size, market status, future development, growth opportunity, challenges, and growth drivers of by analyzing the historical overall data of the considered market segments.

You May Also Like Our Other Top Trending Reports:

For More Information Regarding Other SimiLatest Report Click on The Given Link:https://ksusentinel.com/wp-admin/post-new.php

Contact us:Mr. Abhishek Paliwal5933 NE Win Sivers Drive, #205, Portland,OR 97220 United StatesDirect:+1-971-202-1575Toll Free:+1-800-910-6452E-mail:help@bigmarketresearch.com

More here:
Machine Learning Market 2021 Global Industry Forecasts Analysis, Competitive Landscape and Key Regions Analysis The Manomet Current - The Manomet...