Connect with us

Big Data

If you did not already know




Randomized Block Cubic Newton (RBCN) google

We study the problem of minimizing the sum of three convex functions: a differentiable, twice-differentiable and a non-smooth term in a high dimensional setting. To this effect we propose and analyze a randomized block cubic Newton (RBCN) method, which in each iteration builds a model of the objective function formed as the sum of the natural models of its three components: a linear model with a quadratic regularizer for the differentiable term, a quadratic model with a cubic regularizer for the twice differentiable term, and perfect (proximal) model for the nonsmooth term. Our method in each iteration minimizes the model over a random subset of blocks of the search variable. RBCN is the first algorithm with these properties, generalizing several existing methods, matching the best known bounds in all special cases. We establish ${cal O}(1/epsilon)$, ${cal O}(1/sqrt{epsilon})$ and ${cal O}(log (1/epsilon))$ rates under different assumptions on the component functions. Lastly, we show numerically that our method outperforms the state-of-the-art on a variety of machine learning problems, including cubically regularized least-squares, logistic regression with constraints, and Poisson regression. …

Gaussian Process Latent Variable Alignment Learning google

We present a model that can automatically learn alignments between high-dimensional data in an unsupervised manner. Learning alignments is an ill-constrained problem as there are many different ways of defining a good alignment. Our proposed method casts alignment learning in a framework where both alignment and data are modelled simultaneously. We derive a probabilistic model built on non-parametric priors that allows for flexible warps while at the same time providing means to specify interpretable constraints. We show results on several datasets, including different motion capture sequences and show that the suggested model outperform the classical algorithmic approaches to the alignment task. …

Multimodal Deep Network Embedding (MDNE) google

Network embedding is the process of learning low-dimensional representations for nodes in a network, while preserving node features. Existing studies only leverage network structure information and focus on preserving structural features. However, nodes in real-world networks often have a rich set of attributes providing extra semantic information. It has been demonstrated that both structural and attribute features are important for network analysis tasks. To preserve both features, we investigate the problem of integrating structure and attribute information to perform network embedding and propose a Multimodal Deep Network Embedding (MDNE) method. MDNE captures the non-linear network structures and the complex interactions among structures and attributes, using a deep model consisting of multiple layers of non-linear functions. Since structures and attributes are two different types of information, a multimodal learning method is adopted to pre-process them and help the model to better capture the correlations between node structure and attribute information. We employ both structural proximity and attribute proximity in the loss function to preserve the respective features and the representations are obtained by minimizing the loss function. Results of extensive experiments on four real-world datasets show that the proposed method performs significantly better than baselines on a variety of tasks, which demonstrate the effectiveness and generality of our method. …

Scale-Adaptive Neural Dense Features (SAND Features) google

How do computers and intelligent agents view the world around them? Feature extraction and representation constitutes one the basic building blocks towards answering this question. Traditionally, this has been done with carefully engineered hand-crafted techniques such as HOG, SIFT or ORB. However, there is no “one size fits all” approach that satisfies all requirements. In recent years, the rising popularity of deep learning has resulted in a myriad of end-to-end solutions to many computer vision problems. These approaches, while successful, tend to lack scalability and can’t easily exploit information learned by other systems. Instead, we propose SAND features, a dedicated deep learning solution to feature extraction capable of providing hierarchical context information. This is achieved by employing sparse relative labels indicating relationships of similarity/dissimilarity between image locations. The nature of these labels results in an almost infinite set of dissimilar examples to choose from. We demonstrate how the selection of negative examples during training can be used to modify the feature space and vary it’s properties. To demonstrate the generality of this approach, we apply the proposed features to a multitude of tasks, each requiring different properties. This includes disparity estimation, semantic segmentation, self-localisation and SLAM. In all cases, we show how incorporating SAND features results in better or comparable results to the baseline, whilst requiring little to no additional training. Code can be found at: https://…/SAND_features


Artificial Intelligence

Benefits of Using AI for Facebook Retargeting In 2021




Artificial intelligence has really transformed the state of digital marketing. A growing number of marketers are using AI to connect with customers across various platforms. This includes Facebook.

There are a lot of great reasons to integrate AI technology into your Facebook marketing campaigns. One of the benefits is that you can use retargeting. AI algorithms have made it easier to reach customers that have already engaged with your website. These users might be a lot more likely to convert, which will help you grow your sales and improve your brand image.

AI Can Help Improve Your Facebook Marketing Dramatically with Retargeting

Untapped resources that come from engagement can lead to a better understanding of Facebook retargeting. SAAS SEO agency shared some recommendations that are solid, and a great starting point for any business. To understand Facebook retargeting with AI technology in-depth, take these tips to heart when organizing your resources.

What is Facebook Retargeting?

On average, each Facebook user clicks on ads at least eight times per month. These are considered to be high intent clicks from the biggest advertising platform in the world. Even the most successful marketing and advertising campaigns miss consumers on their first run.

Retargeting uses Facebook campaigns most essential tools to target specific people based on their most relevant data. This is one of the best ways to use data to improve your social media marketing strategies.

The data used is recycled from previous information attached to your old advertising. This includes information from apps, customer files, engagement and offline activity. Anything that has a metric attached to an individual can be used with Facebook retargeting.

How Can It Help Your Business?

There will always be missed opportunities before, during and after a marketing campaign. Reinvesting the data gained from the previous campaign prevents you from starting completely over. Instead of starting from scratch, you’ll gain a clear insight into what gets consumers to cross the finish line at checkout.  Retargeting is meant to be a powerful tool that thrives on previous data that would otherwise go unnoticed.

Retention comes into play, but doesn’t make up the entire story of retargeting on Facebook. You can run a retargeting campaign and only look into new consumers. It’s flexible, and meant to enhance your business based on your specific needs.

The Different Types of Retargeting

The two main types of Facebook retargeting are list-based and pixel-based. Each serves a purpose, with their own specific pros and cons.

Pixel-based retargeting uses JavaScript code to attach a cookie to each unique person that visits your website. After the visitor leaves, the cookie sends its data to your ad provider for a personalized experience. This is the most common type of retargeting used on Facebook, and is often used in other parts of the internet. Microsoft has shown favoritism to pixel-based retargeting by using their own modified version.

List-based retargeting is a limited but fascinating concept. It uses the data you already have on hand to create a specialized list that Facebook uses to show ads. This method works on many of the major social media platforms, but has shown significant advantages on Facebook. Since list-based targeting uses email lists as its base, companies are at the mercy of that particular resource. An outdated or inaccurate email list will lead to low quality retargeting efforts.

When relying on list-based retargeting, a larger email list is not always a guaranteed win for a company.

Upselling and Cross Selling

Even when the customer is happy, proving the value of an upsell is an ongoing process. This led to a rise in cross selling, but was only beneficial to companies that had the resources. As you reconnect with old and new customers, upselling or cross selling becomes part of the closing process.

Both methods are difficult, but become trivial once you have the data to back up your new campaign. Most companies see an increase in profits in a short amount of time. This makes Facebook retargeting a valuable way to test drive upselling and cross selling methods.

Brand Awareness

Brand awareness is the golden goose that all businesses constantly chase. Once you have a notable brand, it becomes the identity of your entire company. Protecting the brand is important, and sometimes entire marketing campaigns are launch to reinvigorate the company image. So, how does Facebook retargeting work its way into this?

Facebook lookalike audiences became a thing when companies wanted to reach new customers with similar interests and habits as their current best customers. Creating a lead that finds this new audience is possible when brand awareness reaches its peak. If you want to keep brand awareness high, then Facebook retargeting does all of the tough work while increasing your reach to new consumers.

Improve Conversions

Conversions are hard to pull off without a specific time investment. All of that goes to waste if you’re not positioning yourself to use previous data to convert customers. No matter how visitors arrive to your website, their presence is proof that there is an interest to purchase a product or service.

If they leave without making a purchase, it’s up to you to figure out why. A lost sale is not the same as losing a customer. Being able to convert that customer into an actual sale is a major strength of retargeting. And even if it’s unsuccessful, you’ll be able to use the additional data to convert another customer.

Influence Buying Decisions

When a consumer becomes firm in their buying decision, then your influence gains a significant bump. At this point your retargeting is directly influencing the buying decisions of individuals or groups. You’ll see a visual representation of this with online feedback and testimonials. All of the positive information provided comes from consumers that are satisfied with the entire sales experience.

Even the negative feedback plays a role, and can serve as the proof you need to reuse data to improve a weak point in your marketing. When a company puts effort into retargeting their ads, they gain monumental increases in customer conversions, ad recognition, clicks, sales and branded searches.

Remarketing Vs. Retargeting

Learn the difference between retargeting and remarketing. Retargeting gains the attention of interested customers that never purchased your products or services. Remarketing leans more towards gaining the attention of inactive or lost customers. Don’t make the mistake of running a retargeting campaign when remarketing would work better. The good news is that the data used from one is still essential for the other. An email list with decent accuracy can be a valuable asset for remarketing or retargeting.

Making the Right Choice

Facebook retargeting should be a priority with how you manage your data collection. Embracing its use will optimize the most important part of your business. Once you get the hang of things, your ROI will reach a whole new level.

Coinsmart. Beste Bitcoin-Börse in Europa

Continue Reading

Big Data

7 Data Collection Tools Every Company Must Have




Businesses need to use their knowledge and data effectively if they want to succeed. This involves collecting plenty of information to ensure your business can make changes and adjustments based on various trends. If you need help collecting and organizing your data, you should get these seven data collection tools to help your company.

Landing Page Tools

Many companies will create landing pages for their businesses. Landing pages are specific spots on your website where people can create accounts, get email newsletters and receive various updates. However, collecting that information on your own will be difficult at times, so you may want a tool to assist you.

Landing page tools will automatically collect that information from your landing pages. On top of that, the tools will automatically add those email addresses and phone numbers to the corresponding newsletters and text services. This makes landing pages ideal if you want to build your text or email channels.

By using this tool, you won’t have to organize the information on your own.

Data Catalog

While data collection remains an important part of the industry, you will also need an effective way to analyze data. This makes data catalog tools an important aspect of your business. As you utilize tools like these, you can figure out what the data means to make informed decisions for your business.

For example, data catalog tools can make your data straightforward and simple to understand if you struggle with analyzing it. It can also help you sort through data to ensure you find information that pertains to you. As you focus on this data and use it, you can assist your business, making data collection tools ideal for most situations.

Referral Programs

Data collection also involves finding more ways for your business to collect useful data. Sure, you could pay to get various emails, but most of those people will get mad at your business for contacting them. This makes referral programs a great option since you can have customers invite their friends to see your business.

Referral programs are great for data collection since your customers can find people for you. From there, the referral tools will collect the data for you, allowing you to send deals and other information to those referrals. Many of these tools will handle referral contacting automatically, so you can boost your sales without putting in any extra work.

Referral programs allow you to work with your customers as you find new ones.

Feedback Collection

When you work with customers, you need to figure out what they think about your business. That way, you can make changes based on what they suggest, allowing you to better appeal to your customers. This makes feedback collection a great way to see what your customers expect from you as a business.

When you focus on feedback collection, you can find out what your customers like and dislike about your business. You can receive feedback in multiple ways, such as through reviews or surveys. Feedback collection tools will automatically gather that information for you, so you can easily view it whenever you want to see what you should change.

SEO Tools

Search engine optimization (SEO) involves your business identifying what keywords people look up online when they want to find information about your business. For example, if you sell computers in Alaska, you will want words like computer, store and Alaska associated with your business. That way, you will appear in those online searches.

SEO involves more than just putting keywords into your website. You also need to identify those keywords, use them effectively and include links that will improve your SEO. Doing this on your own requires tons of effort and time, so you should look for SEO tools to simplify the process and make it easier on you.

If your business has an online presence, you need to look into SEO tools.

Transaction Tracking

When your business makes sales, you can use transaction tracking tools to gather that information for you. For example, you can use these tools to let you know which of your products sell, how often you get sales and when people tend to buy them. This will help you understand your customers’ behavior.

For example, if you notice some of your products tend to sell better than others, you can see what your customers find appealing about those products. This gives you the perfect opportunity to make adjustments to your other products, allowing you to draw in more customers and improve your business overall.

Demographic Tools

Speaking about these tools, it’s also important to understand what types of customers buy from you. Demographics include a variety of points, such as age, where they live, their jobs and lots of other information. By going off the demographics, you can find out what types of people like your products, allowing you to see what you can do to further appeal to them.

For example, if you notice you get lots of younger customers, you can look into trends that they like. Some of your demographics may let you know you should use more technology, though this all depends on the types of customers you get. By using demographic tools, you can quickly find out this information and make decisions based on it.

Demographic tools are key if you want to better understand your customers and what they want from you.


Since data collection plays a major role in the success of companies, you should look into the tools available. Make sure you check these seven tools in particular to assist your business. As you focus on data collection and utilize these tools, you can help your business save time and money.

Coinsmart. Beste Bitcoin-Börse in Europa

Continue Reading


Deepfake detectors and datasets exhibit racial and gender bias, USC study shows




Join Transform 2021 this July 12-16. Register for the AI event of the year.

Some experts have expressed concern that machine learning tools could be used to create deepfakes, or videos that take a person in an existing video and replace them with someone else’s likeness. The fear is that these fakes might be used to do things like sway opinion during an election or implicate a person in a crime. Already, deepfakes have been abused to generate pornographic material of actors and defraud a major energy producer.

Fortunately, efforts are underway to develop automated methods to detect deepfakes. Facebook — along with Amazon  and Microsoft, among others — spearheaded the Deepfake Detection Challenge, which ended last June. The challenge’s launch came after the release of a large corpus of visual deepfakes produced in collaboration with Jigsaw, Google’s internal technology incubator, which was incorporated into a benchmark made freely available to researchers for synthetic video detection system development. More recently, Microsoft launched its own deepfake-combating solution in Video Authenticator, a system that can analyze a still photo or video to provide a score for its level of confidence that the media hasn’t been artificially manipulated.

But according to researchers at the University of Southern California, some of the datasets used to train deepfake detection systems might underrepresent people of a certain gender or with specific skin colors. This bias can be amplified in deepfake detectors, the coauthors say, with some detectors showing up to a 10.7% difference in error rate depending on the racial group.

Biased deepfake detectors

The results, while surprising, are in line with previous research showing that computer vision models are susceptible to harmful, pervasive prejudice. A paper last fall by University of Colorado, Boulder researchers demonstrated that AI from Amazon, Clarifai, Microsoft, and others maintained accuracy rates above 95% for cisgender men and women but misidentified trans men as women 38% of the time. Independent benchmarks of major vendors’ systems by the Gender Shades project and the National Institute of Standards and Technology (NIST) have demonstrated that facial recognition technology exhibits racial and gender bias and have suggested that current facial recognition programs can be wildly inaccurate, misclassifying people upwards of 96% of the time.

The University of Southern California group looked a three deepfake detection models with “proven success in detecting deepfake videos.” All were trained on the FaceForensics++ dataset, which is commonly used for deepfake detectors, as well as corpora including Google’s DeepfakeDetection, CelebDF, and DeeperForensics-1.0.

In a benchmark test, the researchers found that all of the detectors performed worst on videos with darker Black faces, especially male Black faces. Videos with female Asian faces had the highest accuracy, but depending on the dataset, the detectors also performed well on Caucasian (particularly male) and Indian faces. .

According to the researchers, the deepfake detection datasets were “strongly” imbalanced in terms of gender and racial groups, with FaceForensics++ sample videos showing over 58% (mostly white) women compared with 41.7% men. Less than 5% of the real videos showed Black or Indian people, and the datasets contained “irregular swaps,” where a person’s face was swapped onto another person of a different race or gender.

These irregular swaps, while intended to mitigate bias, are in fact to blame for at least a portion of the bias in the detectors, the coauthors hypothesize. Trained on the datasets, the detectors learned correlations between fakeness and, for example, Asian facial features. One corpus used Asian faces as foreground faces swapped onto female Caucasian faces and female Hispanic faces.

“In a real-world scenario, facial profiles of female Asian or female African are 1.5 to 3 times more likely to be mistakenly labeled as fake than profiles of the male Caucasian … The proportion of real subjects mistakenly identified as fake can be much larger for female subjects than male subjects,” the researchers wrote.

Real-world risks

The findings are a stark reminder that even the “best” AI systems aren’t necessarily flawless. As the coauthors note, at least one deepfake detector in the study achieved 90.1% accuracy on a test dataset, a metric that conceals the biases within.

“[U]sing a single performance metrics such as … detection accuracy over the entire dataset is not enough to justify massive commercial rollouts of deepfake detectors,” the researchers wrote. “As deepfakes become more pervasive, there is a growing reliance on automated systems to combat deepfakes. We argue that practitioners should investigate all societal aspects and consequences of these high impact systems.”

The research is especially timely in light of growth in the commercial deepfake video detection market. Amsterdam-based Deeptrace Labs offers a suite of monitoring products that purport to classify deepfakes uploaded on social media, video hosting platforms, and disinformation networks. Dessa has proposed techniques for improving deepfake detectors trained on data sets of manipulated videos. And Truepic raised an $8 million funding round in July 2018 for its video and photo deepfake detection services. In December 2018, the company acquired another deepfake “detection-as-a-service” startup — Fourandsix — whose fake image detector was licensed by DARPA.


VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact. Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Coinsmart. Beste Bitcoin-Börse in Europa

Continue Reading

Big Data

U.S. listing ban on Luokung lifted after judge’s decision




By Karen Freifeld

WASHINGTON (Reuters) – Nasdaq Inc has withdrawn a decision to delist the shares of Luokung Technology Corp, the Chinese mapping technology company said on Thursday, after a U.S. judge suspended an imminent investment ban imposed under the Trump administration.

The ruling and listing news sent shares of the company nearly 20% higher. Luokung issued a news release on Thursday saying Nasdaq notified the company it has withdrawn its delisting letter and shares would continue to trade on the market, not be suspended on May 7. A Nasdaq spokesman declined to comment.

A spokesman for the U.S. Department of Justice did not immediately respond to a request for comment.

Luokung is the second company on a U.S. list of alleged Communist Chinese military companies subject to an investment ban to win a preliminary injunction halting the designation. 

U.S. District Court Judge Rudolph Contreras in Washington issued a similar order in March in favor of Beijing-based smartphone maker Xiaomi Corp.

In granting an injunction in the case brought by Luokung challenging the ban, Contreras said the U.S. Department of Defense’s designation process was flawed.

“Many of the associations the Department of Defense seemed most troubled by – such as Luokung’s purported forays into space systems or its potential future contracts with the Chinese National Geospatial Information Center … do not appear to have materialized, nor are they likely to bear fruit before this case can be decided on the merits,” the judge wrote in his decision.

He added the government has not identified a single technology transfer from Luokung to the People’s Republic of China.

More than 40 companies were added to a list of U.S. companies subject to the investment ban in the waning days of the Trump administration.

(Reporting by Karen Freifeld in New York; Editing by Matthew Lewis)

Image Credit: Reuters

Coinsmart. Beste Bitcoin-Börse in Europa

Continue Reading
Blockchain4 days ago

Ethereum hits $3,000 for the first time, now larger than Bank of America

Blockchain4 days ago

Munger ‘Anti-Bitcoin’ and Buffett ‘Annoyance’ Towards Crypto Industry

Aviation2 days ago

American Airlines Passenger Arrested After Alleged Crew Attack

Blockchain2 days ago

The Reason for Ethereum’s Recent Rally to ATH According to Changpeng Zhao

Blockchain23 hours ago

Chiliz Price Prediction 2021-2025: $1.76 By the End of 2025

Gaming5 days ago

New Pokemon Snap: How To Unlock All Locations | Completion Guide

Blockchain2 days ago

Mining Bitcoin: How to Mine Bitcoin

Blockchain4 days ago

BNY Mellon Regrets Not Owning Stocks of Companies Investing in Bitcoin

Blockchain2 days ago

Mining Bitcoin: How to Mine Bitcoin

Automotive4 days ago

Ford Mach-E Co-Pilot360 driver monitoring system needs an update ASAP

Blockchain5 days ago

Mining Bitcoin: How to Mine Bitcoin

Blockchain5 days ago

Turkey Jails 6 Suspects Connected to the Thodex Fraud Including Two CEO Siblings

Aviation4 days ago

TV Stars Fined After Disorderly Conduct Onboard British Airways

Blockchain4 days ago

Here’s the long-term ROI potential of Ethereum that traders need to be aware of

Blockchain5 days ago

Coinbase to Acquire Crypto Analytics Company Skew

Fintech3 days ago

Talking Fintech: Customer Experience and the Productivity Revolution

AR/VR5 days ago

The dangers of clickbait articles that explore VR

Blockchain5 days ago

A Year Later: Uzbekistan Plans to Lift its Cryptocurrency Ban

Nano Technology4 days ago

Less innocent than it looks: Hydrogen in hybrid perovskites: Researchers identify the defect that limits solar-cell performance

Blockchain4 days ago

Bank of England and UK Parliament get ‘Bitcoin fixes this’ treatment