Tree ensembles, such as random forest and boosted trees, are renowned for their high prediction performance, whereas their interpretability is critically limited. In this paper, we propose a post processing method that improves the model interpretability of tree ensembles. After learning a complex tree ensembles in a standard way, we approximate it by a simpler model that is interpretable for human. To obtain the simpler model, we derive the EM algorithm minimizing the KL divergence from the complex ensemble. A synthetic experiment showed that a complicated tree ensemble was approximated reasonably as interpretable.

Deciding effective and timely preventive measures against complex social problems affecting relatively low income geographies is a difficult challenge. There is a strong need to adopt intelligent automation based solutions with low cost imprints to tackle these problems at larger scales. Starting with the hypothesis that analytical modelling and analysis of social phenomena with high accuracy is in general inherently hard, in this paper we propose design framework to enable data-driven machine learning based adaptive solution approach towards enabling more effective preventive measures. We use survey data collected from a socio-economically backward region of India about adolescent girls to illustrate the design approach.

With the rapid growth of crowdsourcing platforms it has become easy and relatively inexpensive to collect a dataset labeled by multiple annotators in a short time. However due to the lack of control over the quality of the annotators, some abnormal annotators may be affected by position bias which can potentially degrade the quality of the final consensus labels. In this paper we introduce a statistical framework to model and detect annotator's position bias in order to control the false discovery rate (FDR) without a prior knowledge on the amount of biased annotators - the expected fraction of false discoveries among all discoveries being not too high, in order to assure that most of the discoveries are indeed true and replicable. The key technical development relies on some new knockoff filters adapted to our problem and new algorithms based on the Inverse Scale Space dynamics whose discretization is potentially suitable for large scale crowdsourcing data analysis. Our studies are supported by experiments with both simulated examples and real-world data. The proposed framework provides us a useful tool for quantitatively studying annotator's abnormal behavior in crowdsourcing data arising from machine learning, sociology, computer vision, multimedia, etc.

Artificial neural networks are powerful models, which have been widely applied into many aspects of machine translation, such as language modeling and translation modeling. Though notable improvements have been made in these areas, the reordering problem still remains a challenge in statistical machine translations. In this paper, we present a novel neural reordering model that directly models word pairs and their alignment. Further by utilizing LSTM recurrent neural networks, much longer context could be learned for reordering prediction. Experimental results on NIST OpenMT12 Arabic-English and Chinese-English 1000-best rescoring task show that our LSTM neural reordering feature is robust, and achieves significant improvements over various baseline systems. 1 Introduction In statistical machine translation, the language model, translation model, and reordering model are the three most important components.

One startup makes a website that connects high-school students with the ideal college. Another operates a chatbot that can answer your simple medical questions. All have the resources of Microsoft backing them. Nine companies made pitches onstage Thursday night at Showbox SoDo as part of Microsoft Accelerator's third demo day in Seattle. The program selects 10 to 15 companies twice a year to participate in a startup accelerator program that provides resources, Microsoft Azure credits, and -- perhaps most compelling -- introductions to Microsoft's deep pool of customers.

Recovering the planet's natural resources is hard. It's difficult, dangerous, and can be environmentally damaging. Cue an IT revolution, with smart communications, 'extreme Wi-Fi' covering vast deserts, autonomous vehicles that extract vital rocks and minerals, and geofenced employees who receive warnings if they get close to a mine's famously colossal big machinery. There's even a'smart bolt' that creates an underground support structure which is classic Internet of Things. The final goal is the autonomous mine, where humans are completely removed from the mining process.

That's the age of one participant in a recent study looking at ways to take the fingerprints of infants. "The pattern is there at birth," says Anil Jain at Michigan State University in East Lansing. But it is hard to capture. Now Jain and his colleagues are developing a device that could be up to the task. Taking fingerprints from very young children – even newborns – is part of a drive in developing countries to monitor the health of infants, who often lack other forms of identification.

Job Description - Experience and Background: 8-10 years of overall experience 4 years of Solution Designer for case analysis, requirements structuring, data analysis and data modelling to support Enterprise wide decision support systems. Minimum 6 months of experience in Hadoop/Big Data platform Architecture Solution design for Automation based solution specifically using machine or cognitive learning. Areas of Expertise Expertise: in handling huge data loads and optimizing the system to accommodate ever increasing data in the data warehouse/ data store. Rich Data Design & Data Governance Experience Diverse experience in Requirement Capture, Architecture, Solution Design & Delivery Strong Domain experience preferably in Banking, Aviation, Healthcare, Telecom (Either one of them will suffice). Concepts of solution hosted in cloud infrastructure Cognitive or Machine learning based algorithms and implementation Knowledge of concepts like data mining, text mining, data classification, pattern matching, pattern recognition etc.

It is not every day you see a runaway robot causing traffic chaos in a city centre. The robot, named Promobot, was being put through its paces at a research lab in the city of Perm in central Russia's Perm Krai region. The robot, designed to avoid obstacles and to turn around when it reached a boundary, had been left walking around an outside yard. This is the hilarious moment a runaway robot causes traffic chaos in a city centre. The robot - called Promobot - was being put through its paces at a research lab in the city of Perm in central Russia's Perm Krai region Promobot - short for Promotional Robot - is a unique robot created by Russian scientists and is designed to work in customer relations.

Anirudh Koul's grandfather was slowly losing his ability to see. By 2014, he was having a hard time recognizing Koul's face in their weekly Skype calls bridging the vast distance between the Silicon Valley, where Koul is a data scientist at Microsoft, and the elderly man's home in New Delhi. So Koul started reading up on the challenges of vision loss and thinking about how the recent advances in deep learning, a potential-packed area of machine learning, could help give people a new way to recognize what's around them without actually seeing it. That was the modest beginning of Seeing AI. Two years later, Microsoft CEO Satya Nadella introduced the budding technology to thundering applause at this year's Build conference.