DESCRIPTION Are you obsessed with solving challenging problems? Do you have exceptional software engineering skills? Do you think outside of the box and challenge the status quo? Are you constantly looking for ways to improve your skills, your software, and your organization? Our Machine Learning Forecasting team is looking for Software Development Engineers (SDEs) to work on team building a new demand forecasting system in partnership with our Machine Learning team in Berlin.

Gensim is deep learning toolkit implemented in python programming language intended for handling large text collections, using efficient algorithms.

Today at ICML 2016, NVIDIA announced its latest Deep Learning SDK updates, including DIGITS 4, cuDNN 5.1 (CUDA Deep Neural Network Library) and the new GPU Inference Engine. NVIDIA GPU Inference Engine (GIE) is a high-performance deep learning inference solution for production environments. Power efficiency and speed of response are two key metrics for deployed deep learning applications, because they directly affect the user experience and the cost of the service provided. GIE automatically optimizes trained neural networks for run-time performance, delivering up to 16x higher performance per watt on a Tesla M4 GPU compared to the CPU-only systems commonly used for inference today. Figure 1 shows GIE inference performance per watt of the relatively complex GoogLeNet running on a Tesla M4. GIE can deliver 20 Images/s/Watt on the simpler AlexNet benchmark.

As the automotive ecosystem pushes forward with evermore-connected vehicles, new solutions – from over-the-air software updates to connected car services – are emerging from vendors to manage the vast amounts of data that accompany that connectivity. Airbiquity recently announced a software and data management solution specifically for the automotive industry to manage large-scale data collection and software updates to vehicles. SAP introduced its cloud-based Vehicle Insights tool that leverages its HANA big data platform to analyze telematics information as well as other existing business data and external data, with the aim of better integrating connected vehicles into business processes. Last week, IBM said it will be partnering with Local Motors of Maryland on a small, autonomous bus named Olli that incorporates IBM's artificial intelligence and analytics software, Watson, in its first vehicle-based "internet of things" solution – with the big data capabilities focused on passenger interactions rather than the vehicle's self-driving capabilities. Gartner has predicted that by 2020, connected car services will generate almost 40 billion in revenue annually, driven by a combination of infotainmet, navigation, fleet management, traffic management, remote diagnostics and automotive crash notification, among others.

Today, I'm going to explain in plain English the top 10 most influential data mining algorithms as voted on by 3 separate panels in this survey paper. Once you know what they are, how they work, what they do and where you can find them, my hope is you'll have this blog post as a springboard to learn even more about data mining. In order to do this, C4.5 is given a set of data representing things that are already classified. A classifier is a tool in data mining that takes a bunch of data representing things we want to classify and attempts to predict which class the new data belongs to. Sure, suppose a dataset contains a bunch of patients.

Predicting what will happen in the future is challenging. Researchers from MIT's Computer Science and Artificial Intelligence Laboratory developed an algorithm that can predict whether two individuals will hug, kiss, shake hands or slap five in the next scene. Using a Tesla K40 GPU with the cuDNN-accelerated Caffe deep learning framework, the researchers trained their network on 600 hours of prime-time television shows including The Office and Desperate Housewives. When predicting which of the four actions the person would perform one second later, the algorithm correctly predicted the action more than 43 percent of the time – and humans who have been watching TV for years were only able to predict the next action with 71 percent accuracy. In their second study, the algorithm was shown frames from a video and asked it to predict what object will appear five seconds later.

HOW HAS ARTIFICIAL intelligence, associated with hubris and disappointment since its earliest days, suddenly become the hottest field in technology? The term was coined in a research proposal written in 1956 which suggested that significant progress could be made in getting machines to "solve the kinds of problems now reserved for humans…if a carefully selected group of scientists work on it together for a summer". That proved to be wildly overoptimistic, to say the least, and despite occasional bursts of progress, AI became known for promising much more than it could deliver. Researchers mostly ended up avoiding the term, preferring to talk instead about "expert systems" or "neural networks". The rehabilitation of "AI", and the current excitement about the field, can be traced back to 2012 and an online contest called the ImageNet Challenge.

One difficulty faced by artificial intelligence is predicting what humans are going to do next. To help solve that problem, researched have trained an algorithm by making it binge-watch TV. Computer vision experts from MIT's Computer Science and Artificial Intelligence Lab (CSAIL) made an algorithm watch 600 hours of TV shows including Ugly Betty, Scrubs, The Big Bang Theory, The Office (US) and more. In each of the clips, taken from YouTube, humans were performing tasks and interacting with each other. After analysing the videos, the AI was then made to watch a clip it hadn't seen before and predict what would happen.

The ability to make continual, accurate decisions based on evolving data is key in many of today's data-driven intelligent systems. This tutorial-style talk presents an introduction to the modern study of sequential learning and decision making under uncertainty. The broad objective is to cover modeling frameworks for online prediction and learning, explore algorithms for decision making, and gain an understanding of their performance. Specifically, we will look at multi-armed bandits- models of decision making that capture the explore-vs-exploit tradeoff in learning, regret minimization, non-stochastic or adversarial online learning, and online convex optimization. Time permitting, we will discuss new directions and frontiers in the area of sequential decision making.

In machine learning often a tradeoff must be made between accuracy and intelligibility: the most accurate models usually are not very intelligible (e.g., random forests, boosted trees, and neural nets), and the most intelligible models usually are less accurate (e.g., linear or logistic regression). This tradeoff often limits the accuracy of models that can be applied in mission-critical applications such as healthcare where being able to understand, validate, edit, and trust a learned model is important. We have developed a learning method based on generalized additive models (GAMs) that is often as accurate as full complexity models, but remains as intelligible as linear/logistic regression models. In the talk I'll present two case studies where these high-performance generalized additive models (GA2Ms) are applied to healthcare problems yielding intelligible models with state-of-the-art accuracy. In the pneumonia risk prediction case study, the intelligible model uncovers surprising patterns in the data that previously had prevented complex learned models from going to clinical trial, but because it is intelligible and modular allows these patterns to easily be recognized and removed.