In past posts, I've described how Recurrent Neural Networks (RNNs) can be used to learn patterns in sequences of inputs, and how the idea of unrolling can be used to train them. It turns out that there are some significant limitations to the types of patterns that a typical RNN can learn, due to the way their weight matrices are used. As a result, there has been a lot of interest in a variant of RNNs called Long Short-Term Memory networks (LSTMs). As I'll describe below, LSTMs have more control than typical RNNs over what they remember, which allows them to learn much more complex patterns. Lets start with what I mean by a "typical" RNN.

In a recent project, we found which of 700000 property parcels in Adelaide, Australia, contain swimming pools. We used a combination of crowdsourcing and supervised machine learning in order to harness the inherent ability of humans to identify objects in imagery and the speed of machines, which can perform this task much faster than humans, once trained sufficiently. Our initial approach consisted of training a random forest classifier with a set of crowdsourced labels, then using the machine classifications to present to the crowd only the parcels that were likely to contain swimming pools. Since only a small percentage of the parcels actually contain pools, the efficiency gain of this approach is huge compared to a pure crowdsourcing campaign. At first glance, identifying a pool in a high-resolution satellite image might appear to be a simple task for a human and a machine alike.

The notion of applying deep learning techniques to medical imaging data sets is a fascinating and fast-moving area. In fact, in a recent issue of IEEE's Transactions on Medical Imaging journal, there's a fantastic guest editorial on deep learning in medical imaging, that provides an overview of current approaches, where the field is headed, and what sort of opportunities exist. As such, we pulled out some of our favorite nuggets from this article and summarize/extend upon them in Q&A form, so they're more easily digestible. Most interpretations of medical images are performed by physicians; however, image interpretation by humans is limited due to its subjectivity, large variations across interpreters, and fatigue. One way is via transfer learning, which has been used to overcome the lack of large labeled data sets in medical imaging.

When I first spotted Shift Technology with their focus on fraud detection for insurance, I assumed I would find a venture in Israel (which is known for smarts in finding the bad guys in cyberspace, as we outlined when we went to Israel on our Fintech global tour). So I was surprised to find that Shift Technology is a Paris based venture. There is a lot more tech innovation in France than the image of economic sclerosis would lead you to assume. The next thing that jumps out at you is that they recently closed a 10m Series A round in a tough market from a top tier VC (Accel Partners). So they must be doing something right.

With machine learning at the centre of our business platforms, we've been exploring the future of AI. Like many people, I start most mornings digging myself out of an inbox that often feels like more trouble than it's worth thanks to the seemingly infinite amount of newsletters I receive. And although these newsletters cross verticals as varied as marketing, technology, advertising and the agency world, lately each one has a headline that includes artificial intelligence. Not surprisingly, all of us are curious to know how this growing field of technology is going to infiltrate our professional and personal lives. Across every industry, people are afraid that AI is going to steal jobs.

Founded in 2008, Rocket Fuel Inc (NASDAQ:FUEL) is a marketing company that primarily offers a new approach to promoting products and services via its innovative programmatic marketing platform. Rocket Fuel Inc's unique platform utilizes big data and artificial intelligence algorithms to improve the marketing ROI. Rocket Fuel Inc's proprietary Moment Scoring offers a fully data driven marketing approach that aids the programmatic marketing decisions across channels, devices, and objectives for agencies and marketers. Rocket Fuel Inc's platform offers their clients various marketing services, including Data Management Platform (DMP) and Demand Side Platform (DSP). Customers can use these innovative tools themselves, via their own marketing agencies. Rocket Fuel Inc also offers a managed service themselves.

Thanks to the rise of mobile texting and messaging apps, 'chatbots' or "chatter robots" are causing a stir in the world of customer service. Business Insider UK[i] recently wrote that the most powerful tech companies think chatbots are the next best thing since the iPhone. In fact Facebook now allows businesses to deliver automated customer support, e-commerce guidance, content and interactive experiences through chatbots on Facebook Messenger. The company is rumoured to be launching a "bot store" – a move which could be as revolutionary for technology as when Apple launched the App Store. Chatbots are computer programmes that mimic conversations with people using artificial intelligence.

As parents everywhere will attest, the most frequent and difficult question to answer is without doubt'why?'. From an early age people want to know the cause or driving force of everything and it's often the hardest question to answer. It's certainly a challenge when trying to identify how artificial intelligence (AI) algorithms drive uplift. We know the uplift is there, it's clear to see from quarterly results that entertainment campaigns on average see 151% uplift, retail 135% and technology 96%. But where it becomes more difficult is in establishing the individual factors which have produced this uplift. Each time an impression is about to be served information is pulled from trillions of data points which are stored in a Data Management Platform.

In this post I refer to lyrics of certain bands as being "Metal". I know some people have strong feelings about how genres are defined, and would probably disagree with me about some of the bands I call metal in this post. I call these band "Metal" here for the sake of brevity only, and I apologise in advance. It is all around us, and the rate at which it is produced in written, stored form is only increasing. It is also quite unlike any sort of data I have worked with before. Natural language is made up of sequences of discrete characters arranged into hierarchical groupings: words, sentences and documents, each with both syntactic structure and semantic meaning. Not only is the space of possible strings huge, but the interpretation of a small sections of a document can take on vastly different meanings depending on what context surround it.

Since the birth of the modern era of computing there has been an arms race between CPU microprocessor manufacturers that has pushed computer capabilities even higher, characterised by Moore's Law. This era's computer technology can be characterised as running sophisticated but essentially dumb applications. A new era is beginning that will drive microprocessor manufacturers to support intelligent applications, such as those based on newly emerged deep learning and other machine learning algorithms. Deep learning is the umbrella term for a set of techniques for architecting and training neural networks that in recent years has made huge leaps forward in accuracy. For example, deep learning neural networks are at the root of the most successful technologies for natural language understanding, image recognition, advanced game playing (such as Go), and others.