Real-world complex networks describe connections between objects; in reality, those objects are often endowed with some kind of features. How does the presence or absence of such features interplay with the network link structure? Although the situation here described is truly ubiquitous, there is a limited body of research dealing with large graphs of this kind. Many previous works considered homophily as the only possible transmission mechanism translating node features into links. Other authors, instead, developed more sophisticated models, that are able to handle complex feature interactions, but are unfit to scale to very large networks. We expand on the MGJ model, where interactions between pairs of features can foster or discourage link formation. In this work, we will investigate how to estimate the latent feature-feature interactions in this model. We shall propose two solutions: the first one assumes feature independence and it is essentially based on Naive Bayes; the second one, which relaxes the independence assumption assumption, is based on perceptrons. In fact, we show it is possible to cast the model equation in order to see it as the prediction rule of a perceptron. We analyze how classical results for the perceptrons can be interpreted in this context; then, we define a fast and simple perceptron-like algorithm for this task, which can process $10^8$ links in minutes. We then compare these two techniques, first with synthetic datasets that follows our model, gaining evidence that the Naive independence assumptions are detrimental in practice. Secondly, we consider a real, large-scale citation network where each node (i.e., paper) can be described by different types of characteristics; there, our algorithm can assess how well each set of features can explain the links, and thus finding meaningful latent feature-feature interactions.

When was the last time your insurance company realized you had a life change, like getting married or buying a house or having a baby, and designed a policy just for you? And presented it to you without being asked? Without encroaching on personal privacy, this is the vision of insurance in the future: personalized, affordable and adaptable. Artificial intelligence (AI) can help insurance companies understand their customers in powerful new ways and be proactive -- and competitive -- about serving their needs. You may have heard the terms analytics, advanced analytics, machine learning and AI.

Fast forward to 2017 I have spent 100's of hours working on Deep learning projects and the technology has become more and more accessible due to several advancements in software(ease of usage -- Keras, PyTorch), hardware(GPU becoming commercially viable for someone like me sitting in India -Not still cheap), availability of data, good books and MOOCS. After completing the 3 most popular MOOCS in deep learning from Fast.ai, deeplearning.ai/Coursera In this post I talk about 5 aspects of each course which will help you decide. I came across this course when reading an article in kddnudgets . For the first time I heard about Jeremy Howard, searched about him in Wikipedia and was impressed .

Artificial Intelligence and Machine Learning are terms that are currently filling the air in the tech world and across a range of industries, as plans for massive disruption begin to take shape. Both technologies are set to be central to the future, particularly in critical fields such as cybersecurity, and the financial services. Technology is developing at a gallop, and an ever widening skills gap is causing concern, driving the need for technology to fill in for the lack of human skills. However, it is common that the two terms are haphazardly thrown together under the banner of automation, or simply used together without distinction. Because of this, CBR is setting out to find out the main differences, what they are best applied to, and who is standing out in each of the popular spaces.

When I started learning deep learning I spent two weeks researching. I selected tools, compared cloud services, and researched online courses. In retrospect, I wish I could have built neural networks from day one. That's what this article is set out to do. You don't need any prerequisites.

Google's AlphaGo may have unseated Ke Jie as the Go world champion but it's no smarter than a kindergartner. A study published Saturday showed Google's artificial intelligence technology scored best out of 50 systems that Chinese researchers tested against an AI scale they created, although it's still no smarter than a six year old, CNBC reported Monday. AI systems have developed so quickly that it's been able to act as an assistant, take an exam and even outperform us at strategy games.

I've read more than my fair share of articles about all-things AI, including ones on machine learning and deep learning, and two things are clear: (1) my brain is about to short-circuit faster than a vintage robot, and (2) more often than not, the media uses these terms interchangeably even though they're not the same thing. First of all: know that machine learning and deep learning are related and that both fall under the AI umbrella. Think of the three as concentric circles: deep learning is a type of machine learning, machine learning is part of AI, and AI encompasses the entire field of study. In a nutshell, machine learning is an automated process that uses data and math (algorithms) to uncover ("learn") new information without human intervention. Because the machine continuously "self-learns" (or "self-trains"), humans don't need to write code for each process along the way (huge time saver!).

There is a long-standing, red-hot debate in educational circles about the most effective way to teach kids. Others advocate for inquiry-based learning--where students drive their own learning (pdf) through discovery and exploration, working with peers and developing their own ideas--arguing it results in deeper, and more meaningful learning. The two are sometimes pitted against each other as "sage on a stage" (teacher directed) vs. "guide on the side" (student-led, or inquiry based). Both cite ample evidence to prove the superiority of their method (see here for teacher-directed, and here for inquiry-based). McKinsey applied machine learning to the world's largest student database to try and come up with a more scientific answer.

About 4,000 people listened to Cuban as he kicked off his shoes--literally--and explained how AI will change the game for companies, educators, and future developments. He's also keeping his eyes peeled for smaller companies in machine learning and AI, and already has at least three companies in his investment portfolio. "[Software writing] skill sets won't be nearly as valuable as being able to take a liberal arts education … and applying those [skills] in assisting and developing networks." But in order for the country to advance to that future, AI and robotics need to become core competencies in the U.S., and not just in the business world, Cuban said.

My mother's passion for mathematics has always been to the fore, and she has had great pleasure in teaching maths to students, during her career as a teacher, and discusses its incredible impact on any number of walks of life. This has influenced my own interest in putting statistics, data analytics and insight at the centre of business, and I am always curious to understand trends and rhythms to anticipate improvements. Data analysis and targeting technology has delivered productivity gains and heightened customer relevance during the last 20 years. Amazon has used predictive modelling since its launch in 1998. Tesco achieved a significant competitive advantage with the establishment of Club Card and the 50% Share in Dunnhumby in order to mine its data to segment and target customers.