After mastering 40 human languages, a Swedish startup has turned to dolphins, hoping to use its language-analysis software to unlock the secrets of communication employed by the aquatic mammals. Using technology from artificial intelligence language-analysis company Gavagai AB, researchers from Sweden's KTH Royal Institute of Technology will begin compiling a dolphin-language dictionary. The software will monitor captive bottlenose dolphins at a wildlife park about 90 miles south of Stockholm, the company said in an emailed statement Wednesday. "We hope to be able to understand dolphins with the help of artificial intelligence technology," Jussi Karlgren, an adjunct professor of language technology at KTH and co-founder of Gavagai, said in the statement. "We know that dolphins have a complex communication system, but we don't know what they are talking about yet."

One of the most challenging tasks when adopting Bayesian Networks (BNs) is the one of learning their structure from data. This task is complicated by the huge search space of possible solutions and turned out to be a well-known NP-hard problem and, hence, approximations are required. However, to the best of our knowledge, a quantitative analysis of the performance and characteristics of the different heuristics to solve this problem has never been done before. For this reason, in this work, we provide a detailed study of the different state-of-the-arts methods for structural learning on simulated data considering both BNs with discrete and continuous variables, and with different rates of noise in the data. In particular, we investigate the characteristics of different widespread scores proposed for the inference and the statistical pitfalls within them.

A Siamese Deep Forest (SDF) is proposed in the paper. It is based on the Deep Forest or gcForest proposed by Zhou and Feng and can be viewed as a gcForest modification. It can be also regarded as an alternative to the well-known Siamese neural networks. The SDF uses a modified training set consisting of concatenated pairs of vectors. Moreover, it defines the class distributions in the deep forest as the weighted sum of the tree class probabilities such that the weights are determined in order to reduce distances between similar pairs and to increase them between dissimilar points. We show that the weights can be obtained by solving a quadratic optimization problem. The SDF aims to prevent overfitting which takes place in neural networks when only limited training data are available. The numerical experiments illustrate the proposed distance metric method.

In the recent years, we have witnessed the development of multi-label classification methods which utilize the structure of the label space in a divide and conquer approach to improve classification performance and allow large data sets to be classified efficiently. Yet most of the available data sets have been provided in train/test splits that did not account for maintaining a distribution of higher-order relationships between labels among splits or folds. We present a new approach to stratifying multi-label data for classification purposes based on the iterative stratification approach proposed by Sechidis et. al. in an ECML PKDD 2011 paper. Our method extends the iterative approach to take into account second-order relationships between labels. Obtained results are evaluated using statistical properties of obtained strata as presented by Sechidis. We also propose new statistical measures relevant to second-order quality: label pairs distribution, the percentage of label pairs without positive evidence in folds and label pair - fold pairs that have no positive evidence for the label pair. We verify the impact of new methods on classification performance of Binary Relevance, Label Powerset and a fast greedy community detection based label space partitioning classifier. Random Forests serve as base classifiers. We check the variation of the number of communities obtained per fold, and the stability of their modularity score. Second-Order Iterative Stratification is compared to standard k-fold, label set, and iterative stratification. The proposed approach lowers the variance of classification quality, improves label pair oriented measures and example distribution while maintaining a competitive quality in label-oriented measures. We also witness an increase in stability of network characteristics.

We investigate the 3-architecture Connected Facility Location Problem arising in the design of urban telecommunication access networks. We propose an original optimization model for the problem that includes additional variables and constraints to take into account wireless signal coverage. Since the problem can prove challenging even for modern state-of-the art optimization solvers, we propose to solve it by an original primal heuristic which combines a probabilistic fixing procedure, guided by peculiar Linear Programming relaxations, with an exact MIP heuristic, based on a very large neighborhood search. Computational experiments on a set of realistic instances show that our heuristic can find solutions associated with much lower optimality gaps than a state-of-the-art solver.

Dolphins are known to be highly intelligent creatures, and have even been found to construct'sentences' from patterns of clicks and pulses to communicate with each other. And, using artificial intelligence, researchers are now hoping to figure out what they're talking about. Researchers in Sweden are set to begin creating a dolphin-language dictionary using technology from language-analysis startup Gavagai AB – and, it could one day allow humans to communicate with the animals. The program launched by researchers at KTH Royal Institute of Technology and Gavagai AB plan to monitor captive bottlenose dolphins at a wildlife park. The language-analysis software has already proven capable in 40 human languages. And, it's hoped that the artificial intelligence system can similarly decode the dolphins' 'dictionary.'

British politicians who examined the details of a drone strike which killed a British man in Syria said they were disappointed by the government's lack of transparency during investigations. On August 21, 2015, the UK conducted a drone strike in Raqqa for the first time outside the traditional theatre of war, killing 21-year-old British national Reyaad Khan, a suspected fighter with the Islamic State of Iraq and the Levant (ISIL, also known as ISIS), and two other people. "We are in no doubt that Reyaad Khan posed a very serious threat to the UK," the Intelligence and Security Committee in the UK said in a report on Wednesday. "There is nevertheless a question as to how the threat is quantified and in this instance whether the actions of Khan and his associates amounted to an'armed attack' against the UK or Iraq - which is clearly a subjective assessment," the committee said. "The [government's] failure to provide what we consider to be relevant documents is profoundly disappointing," the report added.

The Drum in partnership with intent marketing specialist, Iotec, has launched a report exploring the application of machine learning to solve commercial challenges. The report, Machine Learning: Empowering the Next Generation of Marketing, examines what implications machine learning has and will have on the marketing world, identifies common misconceptions and sheds light on applicable AI-driven marketing solutions. A growing body of research indicates that machine learning is moving to the top of the marketing agenda. A survey by Demandbase and Wakefield Research revealed that 80% of marketing executives believe that AI will revolutionise marketing over the next five years. But the same survey found that only 26% are confident in their understanding of AI technologies and its application to marketing.

Anki had a huge holiday toy hit with the Cozmo Robot. Cozmo was one of the hardest to find toys in the Holiday 2016 shopping season. Today the robotics and artificial intelligence (A.I.) company, announced Anki Overdrive: Fast & Furious Edition launching in September 2017. Don't Miss: This is How to Find a Nintendo Switch in Stock Anki Overdrive: Fast & Furious Edition will merge the most thrilling elements of the hit robotic battle-racing game driven by A.I. with the adrenaline-fueled world of Fast & Furious. Anki Overdrive: Fast & Furious Edition will be available for pre-order starting May 16 for $169.99 at Anki.com and will hit store shelves in September 2017 in the U.S., Canada, U.K., Germany, and Nordic countries.

Last Friday, the government revealed how the budget for more investment in cutting-edge technology and innovation would be split, with business secretary Greg Clark announcing that robotics and AI will be receiving £93 million as part of the government's £1 billion Industrial Strategy Challenge Fund – impacting a range of sectors, including healthcare. In the healthcare sector, technology has already been used to update patient records, improve care delivery and streamline processes. Artificial intelligence (AI) is increasingly being heralded as a technology to achieve further breakthroughs in this sector. UK consumers are also seeing the advantages of introducing AI into the healthcare sector. Recent research from the enterprise information management company, OpenText, revealed that a the UK public would appreciate quicker diagnoses. This was identified as the biggest benefit for people surveyed, with one in three (33%) UK residents believing robots would reach a decision on their condition much faster.