Mediaboss Marketing

image:Han-Seok Seo view more

Credit: University Relations

Could artificial intelligence be used to assist with the early detection of autism spectrum disorder? Thats a question researchers at the University of Arkansas are trying to answer. But theyre taking an unusual tack.

Han-Seok Seo, an associate professor with a joint appointment in food science and the UA System Division of Agriculture, and Khoa Luu, an assistant professor in computer science and computer engineering, will identify sensory cues from various foods in both neurotypical children and those known to be on the spectrum. Machine learning technology will then be used to analyze biometric data and behavioral responses to those smells and tastes as a way of detecting indicators of autism.

There are a number of behaviors associated with ASD, including difficulties with communication, social interaction or repetitive behaviors. People with ASD are also known to exhibit some abnormal eating behaviors, such as avoidance of some if not many foods, specific mealtime requirements and non-social eating. Food avoidance is particularly concerning, because it can lead to poor nutrition, including vitamin and mineral deficiencies. With that in mind, the duo intend to identify sensory cues from food items that trigger atypical perceptions or behaviors during ingestion. For instance, odors like peppermint, lemons and cloves are known to evoke stronger reactions from those with ASD than those without, possibly triggering increased levels of anger, surprise or disgust.

Seo is an expert in the areas of sensory science, behavioral neuroscience, biometric data and eating behavior. He is organizing and leading this project, including screening and identifying specific sensory cues that can differentiate autistic children from non-autistic children with respect to perception and behavior. Luu isan expert in artificial intelligence with specialties in biometric signal processing, machine learning, deep learning and computer vision. He will develop machine learning algorithms for detecting ASD in children based on unique patterns of perception and behavior in response to specific test-samples.

The duo are in the second year of a three-year, $150,000 grant from the Arkansas Biosciences Institute.

Their ultimate goalis to create an algorithm that exhibits equal or better performance in the early detection of autism in children when compared to traditional diagnostic methods, which require trained healthcare and psychological professionals doing evaluations, longer assessment durations, caregiver-submitted questionnaires and additional medical costs. Ideally, they will be able to validate a lower-cost mechanism to assist with the diagnosis of autism. While their system would not likely be the final word in a diagnosis, it could provide parents with an initial screening tool, ideally eliminating children who are not candidates for ASD while ensuring the most likely candidates pursue a more comprehensive screening process.

Seo said that he became interested in the possibility of using multi-sensory processing to evaluate ASD when two things happened: he began working with a graduate student, Asmita Singh, who had background in working with autistic students, and the birth of his daughter. Like many first-time parents, Seo paid close attention to his newborn baby, anxious that she be healthy. When he noticed she wouldnt make eye contact, he did what most nervous parents do: turned to the internet for an explanation. He learned that avoidance of eye contact was a known characteristic of ASD.

While his child did not end up having ASD, his curiosity was piqued, particularly about the role sensitivities to smell and taste play in ASD. Further conversations with Singh led him to believe fellow anxious parents might benefit from an early detection tool perhaps inexpensively alleviating concerns at the outset. Later conversations with Luu led the pair to believe that if machine learning, developed by his graduate student Xuan-Bac Nguyen, could be used to identify normal reactions to food, it could be taught to recognize atypical responses, as well.

Seo is seeking volunteers 5-14 years old to participate in the study. Both neurotypical children and children already diagnosed with ASD are needed for the study. Participants receive a $150 eGift card for participating and are encouraged to contact Seo athanseok@uark.edu.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.

See the rest here:
Researchers using artificial intelligence to assist with early detection of autism spectrum disorder - EurekAlert

Those who own an outdoor cat or even several might run into the occasional problem of having to let them back in. Due to finding it annoying when having to constantly monitor for when his cat wanted to come inside the house, GitHub user gamename opted for a more automated system.

The solution gamename came up with involves listening to ambient sounds with a single Raspberry Pi and an attached USB microphone. Whenever the locally-running machine learning model detects a meow, it sends a message to an AWS service over the internet where it can then trigger a text to be sent. This has the advantage of limiting false events while simultaneously providing an easy way for the cat to be recognized at the door.

This project started by installing the AWS command-line interface (CLI) onto the Raspberry Pi 4 and then signing in with an account. From here, gamename registered a new IoT device, downloaded the resulting configuration files, and ran the setup script. After quickly updating some security settings, a new function was created that waits for new messages coming from the MQTT service and causes a text message to be sent with the help of the SNS service.

After this plethora of services and configurations had been made to the AWS project, gamename moved onto the next step of testing to see if messages are sent at the right time. His test script simply emulates a positive result by sending the certificates, key, topic, and message to the endpoint, where the user can then watch as the text appears on their phone a bit later.

The Raspberry Pi and microSD card were both placed into an off-the-shelf chassis, which sits just inside the house's entrance. After this, the microphone was connected with the help of two RJ45-to-USB cables that allow the microphone to sit outside inside of a waterproof housing up to 150 feet away.

Running on the Pi is a custom bash script that starts every time the board boots up, and its role is to launch the Python program. This causes the Raspberry Pi to read samples from the microphone and pass them to a TensorFlow audio classifier, which attempts to recognize the sound clip. If the primary noise is a cat, then the AWS API is called in order to publish the message to the MQTT topic. More information about this project can be found here in gamename's GitHub repository.

Read this article:
This Smart Doorbell Responds to Meowing Cats Using Machine Learning and IoT - Hackster.io

Joel T. McKelvey

Joel is vice president of product and partner marketing at Sisu, the AI and ML-powered decision intelligence engine that analyzes data at machine scale. A former product manager at Google and leader of product marketing at Looker, he has an extensive background in data and analytics, including business intelligence, database and data storage, and analytics deployment models.

If youre a data analyst, youve probably been approached by company stakeholders asking you questions like: Why is revenue down? Which customers are most likely to churn? What are my top channels to acquire new customers? Why is my business losing more orders in rural areas?

Data analysts know the answers to these questions lie somewhere within their ever-growing troves of company data. However, stakeholders often dont understand the complexity inherent in answering these questions, particularly when dealing with data at cloud scale. In many cases, answers to important business questions are revealed days or weeks later, slowing down decision-making processes and affecting the businesss bottom line.

According to a recent report from McKinsey:

Many business problems still get solved through traditional approaches and take months or years to resolve. By 2025, nearly all employees [will] naturally and regularly leverage data to support their work. Rather than defaulting to solving problems by developing lengthy, sometimes multiyear, road maps, theyre empowered to ask how innovative data techniques could resolve challenges in hours, days or weeks.

As the modern data stack continues to evolve, the amount of data companies collect continues to increase. This progression in data volume, variety and velocity ushers in a new challenge: combing through all of the available data to generate business value.

A recent Gartner report revealed, The volume and velocity of data and increased complexities in decision-making have become too much for a human being to handle without assistance.

So what is the answer? Putting the power of automation in the hands of data teams.

Data teams are starting to understand that operationalized machine learning-powered analytics can increase efficiency and eliminate rote data science work. The ability to rapidly process cloud-scale data, separating signal from the noise with pre-built and operationalized artificial intelligence/machine learning tools, is a necessity for analysts in todays complicated data-rich era.

Analysts today are bottlenecked by tools that mandate the time-consuming manual analysis of data. Analysts spend days or weeks manually defining and testing hypotheses to identify the causal factors behind changing business performance. But its not the analyst who is at fault. Most analytics tools allow analysts to pivot dimensions against each other and to explore data and are very useful, but even as a seasoned analyst, youre probably only able to test one or two hypotheses per minute.

When comprehensive, accurate analysis requires testing millions or billions of hypotheses, analysts often cant respond in time to business needs. Further, analyst teams are forced by limited resources to prioritize the questions they answer, as they simply dont have the resources to support all the decision-makers who require support.

Despite the challenges of scale and complexity, most organizations are able to understand changes that are happening within their data through traditional BI tools. However, most dont realize manually tracking what happens to metrics is only the first step in the decision-making process.

Strong data-backed decision-making doesnt stop after learning business status (what is going on) because what doesnt tell us why it is happening or how to go about addressing it (what next). Understanding and communicating why and what next is the sweet spot where human input and machine automation come together to drive value from data. Effectively answering what, why and what next relies on new ways of tying together people, processes and advanced technology into a single system: decision intelligence.

People are the keystone in the puzzle of getting value from data, particularly complex cloud-scale data. Machine learning and automated delivery of important facts are also only one part of the puzzle. A human has to take these facts and explore them against what is currently happening in the business.

Putting the power of machine learning in the hands of analysts by deploying decision intelligence tools allows them to quickly, proactively and automatically iterate upon the what, why, and what next to quickly and efficiently determine how to prevent issues like customer churn or take advantage of opportunities like the best channels to acquire new customers.

Tools like the Sisu Decision Intelligence Engine help companies wherever their data is housed, whether it be a warehouse or metrics store, and answer those tough questions on what, why, and what next to optimize business performance.

If your organization is looking for a more efficient way to leverage its data to drive business impact, it is important to remember that adding a decision intelligence tool to your tech stack does not replace your BI tools or data science team. In fact, decision intelligence helps data science teams by making them more efficient and helps data scientists focus on the most relevant areas of their data.

By automating the combing through all of a companys trillions of data points to surface insights, data scientists are freed up for more strategic, less repetitive work. A decision intelligence tool is meant to supplement data efforts by performing hypothesis testing at a massive scale and at a fraction of the time of humans alone.

Decision intelligence augments existing BI and data science processes to improve efficiency and feed teams with insights that matter the most to present what, why, and what next through existing interfaces.

Decision intelligence helps organizations drive business outcomes by augmenting people with advanced analytics capabilities integrated directly into decision-making and operational processes. At Sisu, we believe that decision intelligence is what marries people, process and technology together, extracts the most value from data and drives transformational business change.

Feature image via Nappy.

Read more from the original source:
Combining Analyst and Machine Power to Drive Business Results - thenewstack.io

SEATTLE, Aug. 19, 2022 /PRNewswire/ -- The 15th annual Artificial General Intelligence (AGI) Conference opens today at Seattle's Crocodile Venue. Running from August 19-22, the AGI conference event includes in-person events, live streaming, and fee-based video accessand features a diverse set of presentations from accomplished leaders in AI research.

As the AGI community convenes, it continues to promote efforts to democratize AI access and benefits. To that end, several AGI-22 presentations will officially launch DRLearneran open source project to broaden AI access and innovation by distributing AI/Machine Learning code that rivals or exceeds human intelligence across a diverse set of widely acknowledged benchmarks. (Within the AI research community these Arcade Learning Environment [ALE] benchmark tests are widely accepted as a proxy for situational intelligence.)

"Until now, tools at this level in 'Deep Reinforcement Learning' have been available only to the largest corporations and R&D labs," said project lead Chris Poulin. "But with the open-source release of the DRLearner code, we are helping democratize access to state-of-the-art machine learning tools of high-performance reinforcement learning," continued Poulin.

Ben Goertzel, Chairman of the AGI Society and AGI Conference Series, contextualized DRLearner as well-aligned with the goals of the AGI conference. "Democratizing AI has long been a central mission, both for me and for many colleagues. With AGI-22 we push this mission forward by fostering diversity in AGI architectures and approaches, beyond the narrower scope currently getting most of the focus in the Big Tech world," Goertzel said.

DRLearner project presentations include:

"Open Source Deep Reinforcement Learning" General Interest Keynote presented by Chris Poulin, Project Lead. (Journalists Note: Poulin's initial keynote is scheduled for Sunday, August 21. On this day the AGI-22 Conference is open to the general public.)

"Open Source Deep Reinforcement Learning: Deep Dive" Technical Keynote by Chris Poulin and co-principal author Phil Tabor. (Monday, August 22)

"Demo of Open Source DRLearner Tool" Code Demo by co-author Dzvinka Yarish (Monday, August 22)

Story continues

Poulin also noted the importance of managing expectations on the benefits on what DRLearner will, and will not, provide in its initial Beta release: "Fully implementing this state-of-the-art ML capability requires considerable computational power on the cloud, so we advise implementors to maintain realistic expectations regarding any deployment". DRLearner's benefits could be substantial, however, for the numerous organizations who have substantial computing budgets: analytical insights, expanded research capability, and perhaps a competitive advantage. "And for those whose professional lives are focused on AGI, this is an exciting time, as DRLearner can enhance their neural network training efforts" Poulin said.

Drawing on his working experience with both US and Ukrainian computer scientists and software developers, Poulin assembled an international team of expert developers to complete the open-source project. (See more about 'DRLearner's International Dev Team' below.)

A final noteworthy addition, is that the work of Poulin et al was advised by Adria Puigdomenech Badia of DeepMind. "DRLearner provides a great implementation of reinforcement learning algorithms, specifically including the curiosity approach that we had pioneered at DeepMind," said Puigdomenech Badia. Poulin likewise had high praise for the DeepMind's prior "Agent 57" achievement: "Agent 57 was one of a limited number of implementations (at Deep Mind) that consistently beat human benchmarks. And due to the elegant simplicity of its particular design, and help of Adria, it was the best candidate to inspire our software implementation," Poulin said.

ON ARTIFICIAL GENERAL INTELLIGENCE & THE AGI CONFERENCE GOALS

The original goal of the AI field was the construction of "thinking machines"computer systems with human-like general intelligence. Given the difficulty of that challenge, however, AI researchers in recent decades have focused instead on "narrow AI"systems displaying intelligence regarding specific, highly constrained tasks. But the AGI conference series never gave up on this field's ambitious vision; and throughout its fifteen-year existence AGI has promoted the resurgence of broader research on "artificial intelligence"in the original sense of that term.

And in recent years more and more researchers have recognized the necessity and feasibility of returning to the original goals of the field. Increasingly, there is a call for a transition back to confronting the more difficult issues of "human level intelligence" and "artificial general intelligence (AGI)." AGI leaders are committed to continuing the organization's longstanding leadership roleby encouraging and exploring interdisciplinary research based on different understandings of intelligence.

Today, the AGI conference remains the only major conference series devoted wholly and specifically to the creation of AI systems possessing general intelligence at the human level, and ultimately beyond. By convening AI/ML researchers for presentations and discussions, AGI conferences accelerate progress toward our common general intelligence goal.

About the AGI-22 Conference: visit https://agi-conf.org/2022/

About the DRLearner Project: visit http://www.drlearner.org

About Chris Poulin: Poulin specializes in real-time prediction frameworks at Patterns and Predictions, a leading firm in predictive analytics and scalable machine learning. Poulin is also an Advisor at Singularity NET & Singularity DAO. Previously at Microsoft, Poulin was a subject-matter-expert (senior director) in machine learning and data science. He also served as Director & Principal Investigator of the Durkheim Project, a DARPA-sponsored nonprofit collaboration with the U.S. Veterans Administration. At Dartmouth College, Poulin was co-director of the Dartmouth Meta-learning Working Group, and IARPA-sponsored project focused on large-scale machine learning. He also has lectured on artificial intelligence and big data at the U.S. Naval War College. Poulin is co-author of the book Artificial Intelligence in Behavioral and Mental Health (Elsevier, 2015). Chris Poulin's LinkedIn Profile

About Ben Goertzel: Chairman of the AGI Society and AGI Conference Series, Goetzel is CEO of SingularityNET, which brings AI and blockchain together to create a decentralized open market for AIs. SingularityNET is a medium for AGI creation and emergence, a way to roll out superior AI-as-a-service to vertical markets, and a vehicle for enabling public contributions toand benefits fromartificial intelligence. In addition to AGI, Goetzel's passions include life extension biology, philosophy of mind, psi, consciousness, complex systems, improvisational music, experimental fiction, theoretical physics, and metaphysics. For general links to various of his pursuits present and past, see the Goetzel.org website. Ben Goetzel's LinkedIn Profile

About Adria Puigdomenech Badia: For the past seven years Badia has been at DeepMind, where he has specialized in the development of deep reinforcement learning algorithms. Examples of this include 'Asynchronous Methods for reinforcement learning' where he and Vlad Mnih (DeepMind) proposed A3C - 'Neural episodic control'. Badia's recent projects include 'Never Give Up' and 'Agent57' algorithms, addressing one of the most challenging problems of RL: the exploration problem.

DRLearner's International Dev Team:

Chris Poulin (Project Lead-US)Phil Tabor (Co-Lead-US)Dzvinka Yarish (Ukraine)Ostap Viniavskyi (Ukraine)Oleksandr Buiko (Ukraine)Yuriy Pryyma (Ukraine)Mariana Temnyk (Ukraine)Volodymyr Karpiv (Ukraine) Mykola Maksymenko (Advisor-Ukraine)Iurii Milovanov (Advisor-Ukraine)

For media inquiries about the DRLearner project, please contact:

Gregory PetersonArchetype Communicationsgpeterson@archetypecommunications.com

For general inquiries about the AGI-22 Conference, please contact:

Jenny CorlettApril Sixsingularitynet@aprilsix.com

SOURCE drlearner.org

Read more:
At Artificial General Intelligence (AGI) Conference, DRLearner is Released as Open-Source Code -- Democratizing Public Access to State-of-the-Art...

The average U.S. high school class has 30 students, according to research from theNational Council on Teacher Quality, and while each student learns in their own way, practice and specific feedback are repeatedly shown to be effective in modern classrooms. With interactive tools likepractice sets, students can receive one-to-one feedback and support without ever leaving an assignment. This saves the educators time, while also providing insight into students learning processes and patterns.

Achieving both aims at once sounds like a tall order, but adaptive learning technologies helpto do just that. Adaptive learning, a model where students are given customized resources and activities to support their unique learning needs, has been around for decades. However, applying advancing artificial intelligence technology opens up a new set of possibilities to transform the future of school into a personal learning experience.

Google for Educationrecently expanded its suite of adaptive learning tools using artificial intelligence, machine learning and user-friendly design to bring robust capabilities into the classroom.

For educators, adaptive learning technologies help boost instruction, reduce administrative burdens and deliver actionable insights into students progress. More time for planning and catch-up work would help alleviate teachers stress, according to anEdWeek Research Center survey.

For students, adaptive learning tech can deepen comprehension of instructional concepts and help them achieve their personal potential. Through interactive lessons and assignments, real-time feedback and just-in-time support, students can advance through lessons in ways that help increase the likelihood of success.

LEARN MORE:Discover how Google for Education supports students and teachers with CDWG.

When a student grasps a new concept, it can create a magical moment where they suddenly get it, says Shantanu Sinha, vice president and general manager of Google for Education. Ensuring that students get access to the right content or material at the right time is a critical part of making this happen.

By prioritizing students individual learning needs and adapting instruction accordingly, personal learning delivers various benefits, from a well-rounded learning experience to increased productivity, according toeducational publisher Pearson.

Practice setsoffer immediate, personal feedback, which is one of the best ways to keep students engaged. When students are on the right track, fast feedback helps them build confidence and celebrate small wins. When students struggle, real-time feedback helps to ensure they truly understand the material before advancing through a lesson.

Making these experiences interactive can dramatically improve the feedback loop for the student, says Sinha. The ability to see their progress and accuracy when working on an assignment, as well as helpful additional content, can guide students and help them learn.

For instance, Google for Education practice sets use AI to deliver encouragement and support the moment students need them. This includes hints, pop-up messages, video lessons and other resources.

Click the bannerbelow to find resources from CDW to digitally transform your classroom.

With practice sets, teachers can build interactive assignments from existing content, and the software automatically customizes support for students. Practice sets also grade assignments automatically, with the AI recognizing equivalent answers and identifying where students go off track. All these capabilities help teachers extend their reach and maximize their time.

Practice sets also leverage AI to provide an overview of class performance and indicate trends. If several students are having trouble with a concept, teachers can see patterns and adjust quickly without manually sorting through students results.

AI-driven technology opens new opportunities for flexible teaching and learning options. OnChromebooks, for instance, teachers can use Screencast to record video lessons. AI transcribes the spoken lessons into text, allowing students to translate those transcripts into dozens of languages.

Googles adaptive learning tools have built-in, best-in-class security and privacy to protect students and educators personal information. Transparency, multilayered safeguards and continuous updates to ensure compliance with new legislation and best practices are central to delivering adaptive instruction that is secure.

Educators can see and manage security settings on Chromebooks andGoogle Workspacefor Education. IT administrators have visibility via Google for Educations Admin Console.

LEARN MORE:How can a Google Workspace for Education audit benefit your K12 district?

Screencast onChrome OSand practice sets inGoogle Classroomare Googles newest offerings in adaptive learning. Other useful tools include:

As adaptive learning technology continues to evolve, it has the potential to transform the learning experience and help teachers better meet students where they are in the learning journey. When the right technology is applied to teaching and learning, teachers and students can go further, faster.

Brought to you by:

Here is the original post:
Google's Adaptive Learning Technologies Help Amplify Educators' Instruction - EdTech Magazine: Focus on K-12