Leveraging Artificial Intelligence to Create Impressive Products for Disabled People

Technology is an excellent way to enhance the lives of people with disabilities. With the advent of artificial intelligence, several avenues of research have opened up that focus on enhancing the lives of people with impairment.

For instance, Facebook has designed an AI tool that can help the blind “see” again. This AI model explains the images on the Facebook feed of a blind person, so the person using the screen reader gets an idea of what is going on in the picture. This means people with visual impairment no longer have to hear a screen reader say “Photo” by “John Doe.” Google’s ‘Look to Speak’ app uses machine learning and computer vision to allow users to control their devices with their eyes

Similarly, OrCam, a Jerusalem-based company, has developed an AI-based called OrCam Read. This handheld device can read full pages or screens of text aloud from any printed or digital surface, including newspapers, books, product labels, and computers and smartphones. Through this device, OrCam aims to help people with reading challenges, such as dyslexia, mild to moderate vision loss, reading fatigue, as well as for those who read large volumes of text.

Even company giants like Microsoft have started a five-year program called ‘AI for Accessibility,’ with an investment of US$25 million, aiming to put AI in the hands of developers to make the world more accessible by providing AI solutions for the specially-abled. Artificial intelligence not only assists people with physical disabilities but is also helping people struggling with learning problems and mental health issues. E.g., Microsoft’s Windows Hello uses biometric login, i.e., fingerprint, face, or iris, which can work for people with physical disabilities or those with dyslexia who might struggle to remember passwords. AI chatbots like Woebot and Wysa are ensuring the availability of consultation for mental health woes, beyond the therapist hours 24/7.

Meanwhile, people suffering from epilepsy can have seizures from blinking lights and animations. This is why accessiBe, a web accessibility platform enables epileptic users to disable various types of animation, such as GIFs and videos so that they can browse the web without complications. Voiceitt is an app for people with speech impediments, including both those who need it temporarily after strokes and brain injuries, and those with more long-term conditions like cerebral palsy, Parkinson’s, and Down’s syndrome. The app uses machine learning to pick up speakers’ unique speech patterns, recognize any mispronunciations, and rectify them before creating an audio or text output. Livio AI, developed by Starkey, an AI medical device company, is a hearing aid that will enhance the hearing experience by quieting all the external noise from the environment and tracking health-related data to enable patients to seek help during emergencies.

Thanks to artificial intelligence, autonomous vehicles also promise to provide people with disabilities more mobility than ever before. Once the self-driving vehicles are fully integrated into society, they can be a resourceful asset for people with different disabilities, including motor impairment. These people would no longer be dependent on other people or public transport.

Further, most of the existing testing methods are highly ineffective at pinpointing learning disabilities like dyslexia or dyscalculia. Artificial Intelligence can help teachers and healthcare professionals diagnose early signs of such conditions and help the students accordingly. For instance, Australian startup Dystech has developed a screening app for early detection of such learning disorders.

Built on Amazon Web Services (AWS), Dystech employs artificial intelligence and machine learning to screen test if the user has dyslexia or dysgraphia. For the former, the app uses datasets of audio recording from both dyslexic and non-dyslexic adults and children to train the AI and relies on users reading aloud words that appear on the screen while being recorded using their smart device during assessment . And for dysgraphia it uses a photo of a handwritten text for screening. After subjected to a 10-minute screening test, app informs users about their likelihood of having dyslexia or dysgraphia.

The post HOW IS ARTIFICIAL INTELLIGENCE TRANSFORMING THE LIVES OF PEOPLE WITH DISABILITIES? appeared first on Artificial Intelligence.

Years ago, Google struggled with how to pitch its cloud offerings. Back in 2017 I suggested that the company should help mainstream enterprises to “run like Google,” but in a conversation with a senior Google Cloud product executive, he suggested that the company shied away from this approach.

The concern? That maybe mainstream enterprises didn’t share Google’s needs, or maybe Google would simply intimidate them.

For the mere mortals that run IT within such mainstream enterprises (read: almost everyone), fear not. It turns out there are many things that Google might do that make no sense for your own IT needs.

Just ask Colm MacCárthaigh, AWS engineer and one of the authors of the Apache HTTP Server, who asked for “examples of technical things that don’t make sense for everyone just because Amazon, Google, Microsoft, Facebook” do them. The answers—excessive uptime guarantees, site reliability engineering, microservices, and mono-repos among the highlights—are instructive.

Excessive uptime guarantees

“Five or five-plus nines availability guarantees,” says Pete Ehlke. “Outside of medicine and 911 call centres, I can’t think of anything shy of FAANG [Facebook, Amazon, Apple, Netflix, and Google] scale that actually needs five nines, and the ROI pretty much never works out.”

I remember this one well from the variety of start-ups for which I worked, as well as when I was at Adobe (whose service-level commitments tend not to be five nines, but are arguably higher than necessary). Are you going to be OK if the multi-player game goes down? Yep. What about Office 365 for a few minutes, or even hours? Yes and yes.

Site reliability engineering

A bit of a spin on devops (though it predates the devops movement), SRE (named in multiple replies to MacCárthaigh) came out of Google in 2003, and was designed to infuse engineering with an operational focus. A few core principles guide SRE:

• Embrace risk
• Utilize service level objectives (SLOs)
• Eliminate toil
• Monitor distributed systems
• Leverage automation and embrace simplicity

Or, as Ben Traynor, who developed Google’s SRE practice, describes it:

SRE is fundamentally doing work that has historically been done by an operations team, but using engineers with software expertise and banking on the fact that these engineers are inherently both predisposed to, and have the ability to, substitute automation for human labour. In general, an SRE team is responsible for availability, latency, performance, efficiency, change management, monitoring, emergency response, and capacity planning.

SREs spend much of their time on automation, with the ultimate goal being to automate away their job. They spend considerable time on “operations/on-call duties and developing systems and software that help increase site reliability and performance,” says Silvia Pressard.

This sounds important, and even more so if you equate “site reliability” with “business availability.” But do most companies really need their developers to become operational experts? SRE might be critical at Google or Amazon, but it’s arguably a heavy lift for most enterprises, tasking developers with too much of an operational load for them to manage it successfully.

Microservices architecture

As commentator “Buzzy” tells it, “Definitely microservices. The number of 20-staff-in-total companies I’ve had to talk down from that ledge….” Nor is he the only one to call out microservices as a needless complication for most enterprises. Many of the replies to MacCárthaigh’s tweet mentioned microservices.

The post No, you don’t have to run like Google appeared first on Artificial Intelligence.

Facebook AI Research (FAIR) has announced the release of Opacus, a high-speed library for applying differential privacy techniques when training deep-learning models using the PyTorch framework. Opacus can achieve an order-of-magnitude speedup compared to other privacy libraries.

The library was described on the FAIR blog. Opacus provides an API and implementation of a PrivacyEngine, which attaches directly to the PyTorch optimizer during training. By using hooks in the PyTorch Autograd component, Opacus can efficiently calculate per-sample gradients, a key operation for differential privacy. Training produces a standard PyTorch model which can be deployed without changing existing model-serving code. According to FAIR,

[W]e hope to provide an easier path for researchers and engineers to adopt differential privacy in ML, as well as to accelerate DP research in the field.

Differential privacy (DP) is a mathematical definition of data privacy. The core concept of DP is to add noise to a query operation on a dataset so that removing a single data element from the dataset has a very low probability of altering the results of that query. This probability is called the privacy budget. Each successive query expends part of the total privacy budget of the dataset; once that has happened, further queries cannot be performed while still guaranteeing privacy.

When this concept is applied to machine learning, it is typically applied during the training step, effectively guaranteeing that the model does not learn “too much” about specific input samples. Because most deep-learning frameworks use a training process called stochastic gradient descent (SGD), the privacy-preserving version is called DP-SGD. During the back-propagation step, normal SGD computes a single gradient tensor for an entire input “minibatch”, which is then used to update model parameters. However, DP-SGD requires computing the gradient for the individual samples in the minibatch. The implementation of this step is the key to the speed gains for Opacus.

For computing the individual gradients, Opacus uses an efficient algorithm developed by Ian Goodfellow, inventor of the generative adversarial network (GAN) model. Applying this technique, Opacus computes the gradient for each input sample. Each gradient is clipped to a maximum magnitude, ensuring privacy for outliers in the data. The gradients are aggregated to a single tensor, and noise is added to the result before model parameters are updated. Because each training step constitutes a “query” of the input data, and thus an expenditure of privacy budget, Opacus tracks this, providing real-time monitoring and the option to stop training when the budget is expended.

In developing Opacus, FAIR and the PyTorch team collaborated with OpenMined, an open-source community dedicated to developing privacy techniques for ML and AI. OpenMined had previously contributed to Facebook’s CrypTen, a framework for ML privacy research, and developed its own projects, including a DP library called PySyft and a federated-learning platform called PyGrid. According to FAIR’s blog post, Opacus will now become one of the core dependencies of OpenMined’s libraries. PyTorch’s major competitor, Google’s deep-learning framework TensorFlow, released a DP library in early 2019. However, the library is not compatible with the newer 2.x versions of TensorFlow.

The post Facebook Open-Sources Machine-Learning Privacy Library Opacus appeared first on Artificial Intelligence.

Robotics has always been a fascinating subject, but most people are reluctant to experiment in robotics imagining the complexity of hardware and knowledge of different software associated with it. Facebook AI researcher’s(FAIR) latest research will motivate such people to start working on robots. PyRobot is a framework that helps AI researches and students to get up and running with a robot within a few hours, without specialized knowledge of hardware or of details such as device drivers, control, etc. FAIR is now open-sourcing PyRobot to help others in the AI community.

The robot operating system forms the base for PyRobot. It gives a reliable arrangement of hardware-independent midlevel APIs to control various robots. PyRobot abstracts insights concerning low-level regulators and interprocess correspondence, so AI (ML) specialists and others can essentially concentrate on building significant AI robotics applications.

PyRobot gives an API and high-level code for orders to control automated development, such as path planning, visual SLAM, joint position control, joint speed control, and joint torque control. Facebook plans to work with individuals from the robotics research network to create benchmark informational collections.

The post Facebook AI open-sources PyRobot to accelerate AI robotics research appeared first on Artificial Intelligence.

The concept of Big Data has been around for some time now, as most of the business organizations have started to understand the benefits it brings to the table. Data Analytics is becoming more of a higher priority for businesses, as more data is generated every day over the internet. On the Social media front alone, Facebook generates 4 million Gigabytes of data every day, including 4 million likes every minute. On the other hand, Instagram generates 95 million posts daily, which includes 2 million posts from advertisers.

With such a high magnitude of data generated on a daily basis, it becomes impossible to keep up without some sort of data analytics. Many businesses have understood the strategic importance of Data Analytics and are heavily investing in it. More than 75% of the companies are utilizing data analytics to their advantage at present. Tech giants such as Google and Tesla have broken into new markets, leveraging data analytics.

Advantages Big Data Analytics Offer

Most organizations today have Big Data at their disposal, and understand the need to harness it. Big Data Analytics helps organizations harness the data and use it to identify new opportunities in the market. It leads to efficient operations, smarter business moves, higher profits, and happier customers. Big Data Analytics adds value to businesses in the following ways:

Better decision making

Thanks to Data Analytics, Organizations can better understand their audience and recognize what strategies would be successful in marketing their brand. This is where engagement metrics come into play. Organizations are now capable of knowing their consumers’ opinions, especially on the areas concerning likes and dislikes. Marketing content and campaigns are then tailored to perfection to the target audience.

New products and services

The ability to gauge customer needs and satisfaction through analytics gives businesses the power to offer customers what they want. According to Davenport, companies are creating new products to meet customer needs using Big Data Analysis at a scale never seen before.

Cost reduction

The advent of cloud-based analytics tools has brought significant cost advantages when it comes to storing large amounts of data. Amazon, Microsoft, and Google with their cloud storage services are at the forefront when it comes to reducing costs through judicious use of Data Analytics.

Enhanced customer experience

The fundamentals of business have not changed – meeting the customers’ needs is still the ultimate objective of any organization. Today, there is more information available than ever before. Data mined from websites and social media can be used to form a complete view of customer behavior and patterns. If used correctly, Data Analytics enables businesses to know more about consumers than ever thought possible.

 The Bottom Line

The bottom line is that business enterprises are now focusing on utilizing their data collected from consumers to climb up the ladder of success and make critical strategic decisions based on data-driven insights.

Big Data has opened up a whole new world of possibilities impacting the business landscape in every industry right from SME’s to Fortune 500 companies. Almost all industries like finance, healthcare, education have adopted this technology with IT leading the way. This clearly indicates that big data is moving from an experimental endeavor to a more practical pursuit within the organizations. Not only can businesses improve productivity by shifting from manual to automated processes, but they can also see a significant boost in profitability.

From visualizing consumer behavior to building loyalty and making sound decisions, Big Data Analytics presents itself as a goldmine of opportunities for businesses.

The post Big Data Analytics: A Goldmine of Opportunities appeared first on Artificial Intelligence.

Have you ever thought about an AI-based machine playing poker with you? If your imagination has gone that wild then Facebook is all set to make it a reality with its new general AI framework called Recursive Belief-based Learning (ReBeL) that can even perform better than humans in poker and with little domain knowledge as compared to the previous poker setups made with AI.

With ReBel, Facebook is also going for multi-agent interactions – which means that the general algorithms will soon have the capacity to be deployed on a large scale and for multi-agent settings as well. The potential applications include workings like auction, negotiations, and cybersecurity or the operation of self-driving cars and trucks.

Facebook’s plan of combining reinforcement learning with search for AI model training can lead to some remarkable advancements. This is because Reinforcement Learning is based on agents learning to achieve goals in order to maximize rewards whereas search is basically defined as a process that starts from the plan to the stage of setting the goal.

One such example is of Deepmind’s Alpha Zero that is based on a similar program to deliver state-of-the-art performance in board games like chess, shogi, and Go. However, the combination falls short when it is being applied for games like poker because of imperfect information that can arise as a result of how the situation in the game changes. Actions then take help from probability or the playing strategy.

Hence, proposing a solution to the problem in the form of ReBel, Facebook researchers have now expanded the notion of “game state” while including the agent’s belief which relies on the state they are in while playing – counting the common knowledge and policies of other players as well.

When working, ReBel trains two AI models; one is of a value network and the other is of policy network. There is reinforcement learning happening with search during the self-play which eventually has resulted into a flexible algorithm that now holds the potential to beat human players.

For a high level, ReBel operates with public belief states rather than going for world states. If that has surprised you then public belief states are there to generalize the notion of “state value” in games with imperfect information like Poker. PBS is also more often regarded as a common-knowledge probability distribution over a limited arrangement of possible actions and states, which we sometimes call history as well.

Now in perfect-information games, PBS can be distilled down to histories just like the way it distills down to world states in two-player zero-sum games. Not to forget that a PBS is actually the decisions that a player can and also the outcomes of the possibilities on one hand.

As soon as ReBel starts to work for every new game, it creates a “subgame” in the beginning which is very much similar to the original one, except for the fact that its roots go back to the initial PBS. The algorithm actually wins by repeating the runtime of “equilibrium-finding” algorithm and then take advantage of the trained value network to create estimates on every stage of the iteration. Furthermore, with enforcement learning, the values come out easily and then added back to the network as training examples. The policies in the “subgame” are also added as examples. The process continues to repeat itself until PBS becomes the new subgame root and completes a certain accuracy threshold.

The researchers also benchmarked ReBel, as a part of the experiment, for games of heads-up no-limit Texas hold’em poker, Liar’s Dice, and turn endgame hold’em. They used 128 PCs with eight graphic cards only to generate the stimulated game data and of course place random bets and stack sizes (ranging from 5000 to 25000 chips) to test its abilities.

ReBel was also trained on a game with one of the best heads up poker players in the world Don Kim and the results turned out to be ReBel playing faster than two seconds per hand across 7,500 hands and how it didn’t take more than 5 seconds for any decision. Overall ReBel scored 165 thousandths – which is a pretty good result when compared to the previous poker playing system by the social media giant Libratus that resulted in 147 thousandths.

To prevent cheating, Facebook has decided that they will not release ReBel’s codebase for Poker. The company only open-sourced Liar Dice’s implementation, which according to researchers is easier to understand and adjust.

The post Facebook’s New Algorithm Can Play Poker And Beat Humans At It appeared first on Artificial Intelligence.

A team of researchers from Facebook have recently developed a poker-playing AI that is capable of beating human players in heads-up, no-limit Texas hold’em poker.

Called Recursive Belief-based Learning (ReBel), the general AI framework learns poker faster than any other previous poker-specific AI, using less domain knowledge, and researchers are claiming this with a supporting experiment.

The AI was pitted against Dong Kim, considered one of the best heads-up players in the world, alongside three other top human players as part of a series of trials, and the outcomes are impressive!

Not only did ReBel played at a faster pace than its human opponents (faster than two seconds per hand and taking not more than five seconds to make a decision across 7,500 hands), it achieved an aggregated score of 165 thousandths of a big blind per game, defeating Kim with a standard deviation of 69. ReBel performed better than Facebook’s previous poker AI Libratus which recorded an aggregated score of 147.

ReBel’s Development & Applications

ReBel fixes common problems encountered in previous AIs by operating two AI models representing value and policy. Contrary to how past AI’s were developed, such as DeepMind’s Alpha Zero that combined reinforcement learning and search using AI model training for a number of board games like Shogi, Go, and chess, ReBel is mainly developed on game state concepts.

This method results in the creation of a public belief state which enables the AI to come up with probabilities according to the sequence of actions and game states. During the decision-making process, all relevant aspects are considered, including the overall pot and chips, as well as the possible result of a given hand. Based on that information, ReBel creates a “subgame” and then incorporates reinforcement learning until it reaches the designated accuracy level.

Because ReBel does not rely heavily on specific domain knowledge, it’s application is more general and universal, especially in aspects that involve uncertainties and information that are not always available, such as in the game of poker.

The researchers believe the ReBel framework can be applied in developing techniques that involve interactions between multiple agents, such as self-driving cars, negotiations, auctions, and cybersecurity – areas that are usually associated with imperfect-information multi-agent interactions.

To prevent possible cheating in real-life high-stakes games, Facebook has opted not to release the ReBel codebase for poker

The post Facebook’s New Poker-Playing AI ReBel Performs Better Than Humans appeared first on Artificial Intelligence.

As we all know by now that the internet has become a huge pool of data of millions, nay, billions of users around the world. Amongst other tech giants like Facebook or Instagram, Google is one of the biggest data collectors of all time. Now, the Mountain View-based tech giant has released a new report showing the number of requests for user data from various government agencies.

Now, as the internet contains all kinds of information of users from almost every part of the world, these data are sometimes used by various government agencies. We have seen some mishaps happening with data mining, like in the Facebook-Cambridge Analytica case or the Clearview AI fiasco. So, to maintain transparency, Google released this report to show how many requests they get from the governments and how they handle the disclosure of the data.

Now, keep in mind, the below information is from one of the three parts of Google’s “Transparency Report“. The “global requests for user information” report shows the number of requests from many countries. However, if you’re looking for how many requests for user information enterprises have made to the company, there is the second section. And if you want to know about the number of requests related to the US National Security, there is the third one.

Now, coming to the “global requests for user information” report, Google categorized it into three sections — Requests by reporting period, percentage of requests where some information is disclosed, and diplomatic legal requests. 

Amongst these three sections, the lowest number of requests are in the “diplomatic legal requests” section. This is because diplomatic legal requests are made between governments of different countries under an official treaty. So, the chances of these requests are pretty low. As per the report, India has only 17 diplomatic legal requests under her name over the past 10 years.

Now, going back to the previous sections of the report, the “request by reporting period” section and the “percentage of requests where some information is disclosed” section, both contain interactive charts that show the information in an easy-to-read graphical format.

In the chart of the first section, you can filter the graph by choosing a specific country or show data of all the countries combined and the time period on which the requests are made.

In the second chart, however, you only have the “country” filter and no “time period” filter as it shows the percentage of requests of the countries.

Now, according to Google, when they receive requests like these from the governments, they “carefully review each request to make sure it satisfies applicable laws. If a request asks for too much information, [they] try to narrow it, and in some cases, [they] object to producing any information at all.”

The post Google’s “Transparency Report” Shows Requests for User Information From Governments appeared first on Artificial Intelligence.

During the last decade, social networking sites/apps have become the most important channels of communication.Social networks such as Facebook, Twitter, and Instagram contain a considerable amount of informative data not only about social matters but also about business and marketing.

Of course, if you want to take advantage of this ever-changing space to swing the balance in your brand’s favor, you don’t have a choice unless using big data analytics.In this post, I’m going to describe how social media can be affected by big data analytics and how businesses can make the most out of it.

What is big data?

Big data is any novel technique used to analyze a massive volume of data that is so large it is impossible to process with traditional methods.Big data can handle both structured and unstructured data and help you tackle the problem of processing power. The main purpose of using big data techniques is finding overall patterns, trends, and connections between different variables.This is particularly important for studying data related to human interactions and social behaviors. This is exactly where social media can be influenced by big data, especially for marketing purposes.In fact, the huge, unstructured knowledge available on social media can’t add real value to your marketing strategy. So you need a powerful tool like big data analytics to be able to handle it.

Important data on social media

The stream of social posts, likes, mentions, shares, followers, and many other impressions can clearly prove why big data is important in social media marketing.This is now a must for businesses to collect these tones of information in real-time and analyze it to know how well their social presence is.In fact, every single positive interaction can add to their reputation and any negative feedback can put their efforts at stake.Without the shadow of a doubt, social media marketing analytics, especially for big companies, can’t be impactful without big data. Therefore, many businesses are investing masses of money on big data tools to track real-time consumer behavior across social media.To do this you will have to:

1. Collect the most related data available on social networks
2. Recognize the weight of each data on your target market
3. Convert these data into useful facts and use them in your strategies

Of course, the velocity of your data mining, the volume of the data, and also its variety is of paramount.You’ll need to use the best data technology and analytical tools if you want to leverage big data in marketing effectively.Also, it’s important to consider integrating various social networking sites/apps. Using Facebook, Instagram, Twitter, and LinkedIn will lead to better social interactions and the data from them can represent reality more exactly.

Advantages of big data for digital marketing

A lot of big and small companies are thinking about considerable budgets for big data analytics tools to get ahead of the competition.Here are 5 top benefits a big data analysis approach can bring to your social media campaign:1. Taking care of huge information sourcesAs a social user, you may need to process all data related to your niche that comes from mainstream channels.Analyzing diverse channels is not an easy task and can only be done by using artificial intelligence and big data technology.A lot of business sites allow users to sign up via Google or other main channels. So marketers can collect and analyze data about their niche customers from social networks, browser history, applications, cloud storage, etc.2. Targeting the right audienceIt’s clear that you can’t reach out to all the internet users so you have to narrow down to the most probable group of customers.Thus, social media marketing is all about identifying your target audience. Big data technology has provided marketers with access to insightful data of users’ personal information, photos, favorites, locations, and various kinds of activities.3. Predicting online behaviorsA big data approach can also be used for better decision-making based on previous trends. Data-based businesses are becoming incredibly efficient, as computers can predict the potential choices of customers.In sum, the interests and habits of people can be estimated as they’re changing based on specific overall trends.4. Managing marketing campaignsBig data techniques enable marketers to accurately track the ROI metrics of their social media campaigns.It will provide advertisers with insightful data into how effective a social media campaign has been or can be. Predictive analytical methods greatly help in predicting what products/services consumers want.Tracking consumer behaviors all over social will clear many things about the effectiveness of previous campaigns.This includes media including their engagement and reaction to online advertising. So marketers can optimize their plans for future campaigns to get a higher ROI.5. Identifying fair pricesOne of the biggest problems for marketers is to find reasonable prices for sponsored ads. A lot of different factors are affecting the prices.For example, during the COVID-19 pandemic, a lot of influencers have considerably cut their rates. So, it’s important to track the accepted costs on the internet and make arrangements accordingly.A big data analysis on social media can be helpful to clearly know what prices your competitors or niche customers are agreeing at.

Social media strategy when using big data

If you want to make sure that your big data approach will lead to a higher ROI on social media, you need to have an efficient strategy.Without a strategy, you don’t know what exactly you want from the unstructured data on social channels. Therefore, you’ll waste your time and money without generating considerable leads and boosting sales.A common social media strategy usually comprise the below steps:

• Market research
• Defining SMART goals
• Identifying the audience
• Choosing the right platforms
• Generating relevant content
• Scheduling posts and establishing a social presence
• Engaging with the community
• Influencer marketing
• Analyzing the performance

This strategy should help you achieve certain goals which usually contain below items:

• Drive traffic to your site
• Get higher conversions rates
• Build brand awareness
• Make you appear like a niche leader

Additionally, you’ll need certain metrics to analyze your performance and show you how well you’re achieving these goals.Actually, great use of a big data strategy is to help you know your social performance. You’ll be able to optimize your social techniques based on this information and make the best out of it.

Conclusion

Big data is one of the newest features of information technology and can be of great use to digital marketers. It was tried to summarize the important aspects of big data analytics in social media marketing and the benefits it can bring to your campaign. Remember to take advantage of this technology to optimize your social presence and get ahead of the marketing competition.

The post How Big Data Can Help to Analyze Social Media Performance appeared first on Artificial Intelligence.

In a technical paper published on Arxiv.org this week, researchers at Facebook and Arizona State University lifted the hood on AutoScale, which shares a name with Facebook’s energy-sensitive load balancer. AutoScale, which could theoretically be used by any company were the code to be made publicly available, leverages AI to enable energy-efficient inference on smartphones and other edge devices.

Lots of AI runs on smartphones — in Facebook’s case, the models underpinning 3D Photos and other such features — but it can result in decreased battery life and performance without fine-tuning. Deciding whether AI should run on-device, in the cloud, or on a private cloud is therefore important not only for end users but for the enterprises developing the AI. Datacenters are expensive and require an internet connection; having AutoScale automate deployment decisions could result in substantial cost savings.

For each inference execution, AutoScale observes the current execution rate, including the architectural characteristics of the algorithm and runtime variances (like Wi-Fi, Bluetooth, and LTE signal strength; processor utilization; voltage; frequency scaling; and memory usage). It then selects hardware (processors, graphics cards, and co-processors) that are expected to maximize energy efficiency while satisfying quality of service and inference targets based on a lookup table. (The table contains the accumulated rewards — values that spur on AutoScale’s underlying models to complete goals — of the previous selections.) Next, AutoScale executes inference on the target defined by the selected hardware while observing its result, including energy, latency, and inference accuracy. Based on this and before updating the table, the system calculates a reward indicating how much the hardware selection improved efficiency.

As the researchers explain, AutoScale taps reinforcement learning to learn a policy to select the best action for an isolated state, based on accumulated rewards. Given a processor, for example, the system calculates a reward with a utilization-based model that assumes (1) processor cores consume a variable amount of power; (2) cores spend a certain amount of time in busy and idle states; and (3) energy usage varies among these states. By contrast, when inference is scaled out to a connected system like a datacenter, AutoScale might calculate a reward using a signal strength-based model that accounts for transmission latency and the power consumed by a network.

To validate AutoScale, the coauthors of the paper ran experiments on three smartphones, each of which was measured with a power meter: the Xiaomi Mi 8 Pro, the Samsung Galaxy S10e, and the Motorola Moto X Force. To simulate cloud inference execution, they connected the handsets to a server via Wi-Fi, and they simulated local execution with a Samsung Galaxy Tab S6 tablet connected to the phones through Wi-Fi Direct (a peer-to-peer wireless network).

After training AutoScale by executing inference 100 times (resulting in 64,000 training samples) and compiling and generating 10 executables containing popular AI models, including Google’s MobileBERT (a machine translator) and Inception (an image classifier), the team ran tests in a static setting (with consistent processor, memory usage, and signal strength) and a dynamic setting (with a web browser and music player running in the background and signal inference). Three scenarios were devised for each:

• A non-streaming computer vision test scenario where a model performed inference on a photo from the phones’ cameras.
• A streaming computer vision scenario where a model performed inference on a real-time video from the cameras.
• A translation scenario where translation was performed on a sentence typed by the keyboard.

The team reports that across all scenarios, AutoScale beat baselines while maintaining low latency (less than 50 milliseconds in the non-streaming computer vision scenario and 100 milliseconds in the translation scenario) and high performance (around 30 frames per second in the streaming computer vision scenario). Specifically, it resulted in a 1.6 to 9.8 times energy efficiency improvement while achieving 97.9% prediction accuracy and real-time performance.

Moreover, AutoScale only ever had a memory requirement of 0.4MB, translating to 0.01% of the 3GB RAM capacity of a typical mid-range smartphone. “We demonstrate that AutoScale is a viable solution and will pave the path forward by enabling future work on energy efficiency improvement for DNN edge inference in a variety of realistic execution environment,” the coauthors wrote.

One of the most popular existing techniques — neural ordinary equations (ODEs) —  have an important limitation in that they can’t account for random interactions, meaning that they can’t update the state of a system as random events occur. (Think trades by other people that affect a company’s share price or a virus picked up at a hospital that changes a person’s health status.) The system has to be updated manually on some schedule to account for these, which means that the model isn’t truly mapping to reality.

Neural SDEs have no such limitation. That’s because they represent continuous changes in state as they occur.

As the coauthors of the paper explain, neural SDEs generalize ODEs by adding instantaneous noise to their dynamics. This and other algorithmic tweaks allow tens of thousands of variables (parameters) to be fitted to a neural SDE, making it a fit for modeling things like the motion of molecules in a liquid, allele frequencies in a gene pool, or prices in a market.

In one experiment, the team trained ODE and neural SDE models on a real-world motion capture data set comprising 23 walking sequences partitioned into 15 training, 3 validation, and 4 test sequences. After 400 iterations, they observed improved predictive performance from the neural SDEs compared with the ODEs — the former had a mean squared error of 4.03% versus the ODE’s 5.98% (lower is better).

“Building on the early work of Einstein, these SDEs enable models to represent continuous changes in state as they occur and to do so at scale,” a Vector Institute spokesperson told VentureBeat via email. “Non-neural SDEs are used in finance and health today, but their scale is limited. As mentioned at the top, neural SDEs introduce the new chance to apply AI at scale to large complex financial systems without having to make the big … compromises that have typically been required.”

The post Facebook’s AutoScale decides if AI inference runs on your phone or in the cloud appeared first on Artificial Intelligence.