Venture capital firm Octopus Ventures has raised a £120 million fund to invest in UK technology startups, The Financial Times reports. The London-based investment company, part of UK fund management firm Octopus Group, has reportedly now raised a total of £660 million, fuelled in part by a string of successful UK startup exits. Natural language processing pioneer Evi Technologies, predictive keyboard startup SwiftKey, and AI video startup Magic Pony are all in the Octopus Ventures portfolio, as are companies like LoveFilm and property website Zoopla. "Evi Technologies in Cambridge was sold to Amazon and now powers Alexa, the voice-activated speaker," Alex Macpherson, chief executive of Octopus Ventures, told the FT. "We also backed Swiftkey which went to Microsoft last year [for $250 million; £200 million] and made a seed investment in Magic Pony, which went to Twitter [for $150m; £120 million].

Arthur Renowitzky can't help but command attention as he walks down the street on a sunny autumn morning. A driver lowers her window to flash a smile and a thumbs-up. "You got this," she says. Renowitzky has been paralyzed since 2007 after being shot in the chest for $20 and a fake gold chain. But he can stand and walk, using crutches for balance, when wearing an exoskeleton suit with motorized hips and knees powering his movements.

As a general rule of thumb, the law and government are slow-moving and deliberate. That's really handy for important things that you have to get right, but the trouble is that disruptive technology tends to move much faster. That's bad enough if the disruptive technology you're talking about is (say) the sharing economy, but it's more serious when it's something that Elon Musk once described as "potentially more dangerous than nukes". That thing is artificial intelligence. And while it's hard to feel too threatened when it's your Amazon Echo failing to understand you saying "Play REM" for the tenth time in a row, the threat – potentially – is a real one.

NEC's artificial intelligence platform is mixing data, bus telematics, and human observation to determine whether a bus driver in Singapore is likely to cause an accident in the next three months. NEC, in partnership with Singapore public transport service provider SMRT Corporation, is using artificial intelligence (AI) to prevent bus accidents on the city-state's roads. The Japanese giant is taking historical data from a bus driver's work records, telematics data produced by each bus, and observations made by on-board data scientists to determine whether a driver is likely to cause an accident in the next three months, and intervene before SMRT has to deal with the costly and potentially life-threatening aftermath. Here's what Samsung's latest budget phones may tell us about the Galaxy S9 Addressing the Tech Leaders Forum 2017 on Monday, Mervyn Cheah, head and vice president of NEC Laboratories Singapore, explained that once an at-risk driver is identified, they are sent for further training. Buses on the road, sometimes they do create accidents and actually they will always react when they have an accident, you'll send them for training, he said.

Low paying jobs seem to be more at risk of a robot takeover, but a new developed has suggests that not even highly paid Wall Street jobs are safe. Goldman Sacs has published a job posting seeking a software developer to build a'robot adviser' that provides mass affluent clients with'detailed information on their financial portfolio and analytics'. The move comes as Goldman is looking at ways to broaden its customer base outside the super wealthy, including making deeper inroads into new consumer-focused businesses. Goldman Sacs has published a job posting seeking a software developer to build a platform that gives mass affluent clients'detailed information on their financial portfolio and analytics' According to the job posting, Goldman Sachs wants to leverage'a global technology platform offering an integrated suite of tools and applications to service clients. 'Digital innovations to shape client experience and enable our Institutional and Third-Party Distribution (TPD) salesforce with the tools required to best serve our clients,' the listing reads.

Representing documents is a crucial component in many NLP tasks, for instance predicting aspect ratings in reviews. Previous methods for this task treat documents globally, and do not acknowledge that target categories are often assigned by their authors with generally no indication of the specific sentences that motivate them. To address this issue, we adopt a weakly supervised learning model, which jointly learns to focus on relevant parts of a document according to the context along with a classifier for the target categories. Derived from the weighted multiple-instance regression (MIR) framework, the model learns decomposable document vectors for each individual category and thus overcomes the representational bottleneck in previous methods due to a fixed-length document vector. During prediction, the estimated relevance or saliency weights explicitly capture the contribution of each sentence to the predicted rating, thus offering an explanation of the rating. Our model achieves state-of-the-art performance on multi-aspect sentiment analysis, improving over several baselines. Moreover, the predicted saliency weights are close to human estimates obtained by crowdsourcing, and increase the performance of lexical and topical features for review segmentation and summarization.

Big Data is one of the major challenges of statistical science and has numerous consequences from algorithmic and theoretical viewpoints. Big Data always involve massive data but they also often include online data and data heterogeneity. Recently some statistical methods have been adapted to process Big Data, like linear regression models, clustering methods and bootstrapping schemes. Based on decision trees combined with aggregation and bootstrap ideas, random forests were introduced by Breiman in 2001. They are a powerful nonparametric statistical method allowing to consider in a single and versatile framework regression problems, as well as two-class and multi-class classification problems. Focusing on classification problems, this paper proposes a selective review of available proposals that deal with scaling random forests to Big Data problems. These proposals rely on parallel environments or on online adaptations of random forests. We also describe how related quantities -- such as out-of-bag error and variable importance -- are addressed in these methods. Then, we formulate various remarks for random forests in the Big Data context. Finally, we experiment five variants on two massive datasets (15 and 120 millions of observations), a simulated one as well as real world data. One variant relies on subsampling while three others are related to parallel implementations of random forests and involve either various adaptations of bootstrap to Big Data or to "divide-and-conquer" approaches. The fifth variant relates on online learning of random forests. These numerical experiments lead to highlight the relative performance of the different variants, as well as some of their limitations.

Machine learning techniques are being increasingly used as flexible non-linear fitting and prediction tools in the physical sciences. Fitting functions that exhibit multiple solutions as local minima can be analysed in terms of the corresponding machine learning landscape. Methods to explore and visualise molecular potential energy landscapes can be applied to these machine learning landscapes to gain new insight into the solution space involved in training and the nature of the corresponding predictions. In particular, we can define quantities analogous to molecular structure, thermodynamics, and kinetics, and relate these emergent properties to the structure of the underlying landscape. This Perspective aims to describe these analogies with examples from recent applications, and suggest avenues for new interdisciplinary research.

Simple Temporal Networks (STNs) provide a powerful and general tool for representing conjunctions of maximum delay constraints over ordered pairs of temporal variables. In this paper we introduce Hyper Temporal Networks (HyTNs), a strict generalization of STNs, to overcome the limitation of considering only conjunctions of constraints but maintaining a practical efficiency in the consistency check of the instances. In a Hyper Temporal Network a single temporal hyperarc constraint may be defined as a set of two or more maximum delay constraints which is satisfied when at least one of these delay constraints is satisfied. HyTNs are meant as a light generalization of STNs offering an interesting compromise. On one side, there exist practical pseudo-polynomial time algorithms for checking consistency and computing feasible schedules for HyTNs. On the other side, HyTNs offer a more powerful model accommodating natural constraints that cannot be expressed by STNs like Trigger off exactly delta min before (after) the occurrence of the first (last) event in a set., which are used to represent synchronization events in some process aware information systems/workflow models proposed in the literature.

Predicting business process behaviour is an important aspect of business process management. Motivated by research in natural language processing, this paper describes an application of deep learning with recurrent neural networks to the problem of predicting the next event in a business process. This is both a novel method in process prediction, which has largely relied on explicit process models, and also a novel application of deep learning methods. The approach is evaluated on two real datasets and our results surpass the state-of-the-art in prediction precision.