Mumbai: Smart machines, including cognitive computing, artificial intelligence, intelligent automation, machine learning and deep learning, will enter mainstream adoption by 2021, with 30% adoption by large companies, a Gartner report said on Friday. The report suggests that this represents opportunities to help enterprises assess, select, implement, change and adapt talent, and for IT and business processes for business benefits. "Smart machines will profoundly change the way work is done and how value is created. From dynamic pricing models and fraud detection, to predictive policing and robotics, smart machines have broad applicability in all industries," said Susan Tan, research vice president at Gartner, in a statement. The report added that the opportunity for consulting and system integration (C&SI) services will range from advising enterprises to help them sort through the hype to helping with strategic design, training of the smart machines, deployment and integration to expansion and ongoing refinement.

IBM's Watson supercomputer to fight real-world cyber security Seven months after first announcing that it was teaching its Watson cognitive technology platform to fight cybercrime, IBM Corp. has launched it into the real world, at least in test mode. The Watson for Cyber Security platform has been designed to discover behaviour patterns and evidence of hidden cyber attacks and threats that could otherwise be missed by existing security platforms. It does so by using Watson's ability to reason and learn from unstructured data, including the 80 percent of all data on the Internet that traditional security tools cannot process, including blogs, articles, videos, reports, alerts and other information. The software incorporates capabilities such data mining for outlier detection, graphical presentation tools and techniques for finding connections between related data points in different documents, including the ability to identify warnings of new types of malware from even obscure sources. In the initial beta phase, customers are not being charged for the service. Some 40 organisations signed on for the beta test, including Sun Life Financial, the University of Rochester Medical Center, Avnet, SCANA Corp., Sumitomo Mitsui Banking Corp., California Polytechnic State University, the University of New Brunswick and Smarttech.

Evernote has reversed proposed changes to its privacy policy that would allow employees to read user notes to help train machine learning algorithms. CEO Chris O'Neill said the company had "messed up, in no uncertain terms." The move by the note-taking app follows protests from users, some of whom have threatened to drop the service after the company announced that its policy would change to improve its machine learning capabilities by letting a select number of employees, who would assist with the training of the algorithms, view the private information of its users. The machine learning technologies would make users more productive as they would allow the automation of functions now done manually, like creating to-do lists or putting together travel itineraries, O'Neill had said earlier on Thursday in defense of the proposed changes. Evernote employees would only see random content in snippets to check that the features are working properly but they wouldn't know who it belongs to, and personal information would be masked, he added.

"Artificial Intelligence" is all set to change our life as well as perspective. With digitization on an incredible rise, AI to have a dominating impact on our life. According to Ericsson Consumer Lab's global research activities of over more than 20 years, representing 27 million citizens as well as data from an online survey of advanced internet users in 14 major cities across the world, AI will become a lot smarter in 2017. It has also found that VR will be indistinguishable from physical reality in three years. Following are the ten ways AI will change the world in 2017.

SAN FRANCISCO – The dwarf planet Ceres, an enigmatic rocky body inhabiting the main asteroid belt between Mars and Jupiter, is rich with ice just beneath its dark surface, scientists said on Thursday in research that may shed light on the early history of the solar system. The discovery, reported in a pair of studies published in the journals Science and Nature Astronomy, could bolster fledgling commercial endeavors to mine asteroids for water and other resources for robotic and eventual human expeditions beyond the moon. NASA's Dawn spacecraft has been orbiting Ceres, the largest of thousands of rocky bodies located in the main asteroid belt, since March 2015 following 14-month study of Vesta, the second-largest object in the asteroid belt. The studies show that Ceres is about 10 percent water, now frozen into ice, according to physicist Thomas Prettyman of the Planetary Science Institute in Tucson, Arizona, one of the researchers. Examining the makeup of solar system objects like Ceres provides insight into how the solar system formed.

Artificial intelligence (AI) has officially gone mainstream. Industry research firm Gartner named AI as its number one strategic technology for a second year in a row. The acquisitions race among giants like Google, IBM, Salesforce and Apple to purchase private AI companies keeps heating up -- 2016 alone saw 40 AI-related acquisitions and our own research found that 62% of large enterprises will be using AI-technologies by 2018. Since everyone seems to be talking about AI broadly, we focused our predictions this year on what we see happening with communications and AI. As a leader in this area, we are working with enterprises to close the communication gap between man and machine.

Knowmail, a company building a Personalized Artificial Intelligence platform to help employees communicate and collaborate better, announced it closed a new $3.5 million investment led by CE Ventures with managing partner Tayman Kan also joining the board. This round also includes existing shareholders AfterDox, Plus Ventures, 2B Angels, INE Ventures and notable private investors. The additional funding will be utilized to continue building out Knowmail's functionality and expand its customer base globally. Founded in 2014, Knowmail is led by Haim Senior (CEO), previously executive at EMC, seasoned entrepreneur Oded Avital (COO) and previous Verint Systems ASG director Avi Mandelberg (CTO) in order to, in their own words, "liberate corporate employees from slavery created by information overload". "We are thrilled to lead this latest round of funding for Knowmail and strongly believe in the team and vision."

Kidney exchanges are organized markets where patients swap willing but incompatible donors. In the last decade, kidney exchanges grew from small and regional to large and national---and soon, international. This growth results in more lives saved, but exacerbates the empirical hardness of the $\mathcal{NP}$-complete problem of optimally matching patients to donors. State-of-the-art matching engines use integer programming techniques to clear fielded kidney exchanges, but these methods must be tailored to specific models and objective functions, and may fail to scale to larger exchanges. In this paper, we observe that if the kidney exchange compatibility graph can be encoded by a constant number of patient and donor attributes, the clearing problem is solvable in polynomial time. We give necessary and sufficient conditions for losslessly shrinking the representation of an arbitrary compatibility graph. Then, using real compatibility graphs from the UNOS nationwide kidney exchange, we show how many attributes are needed to encode real compatibility graphs. The experiments show that, indeed, small numbers of attributes suffice.

In this short report, we discuss an approach to estimating causal graphs in which indicators of causal influence between variables are treated as labels in a machine learning formulation. Available data on the variables of interest are used as "inputs" to estimate the labels. We frame the problem as one of semi-supervised learning: available interventional data or background knowledge provide labels on some edges in the graph and the remaining edges are treated as unlabelled objects. To illustrate the key ideas, we consider a simple approach to feature construction (rooted in bivariate kernel density estimation) and embed this within a semi-supervised manifold framework. Results on yeast knockout data demonstrate that the proposed approach can identify causal relationships as validated by unseen interventional experiments. An advantage of the formulation we propose is that by reframing causal discovery as semi-supervised learning, it allows a range of data-driven approaches to be brought to bear on causal discovery, without demanding specification of full probability models or explicit models of underlying mechanisms.

Directed graphs occur throughout statistical modeling of networks, and exchangeability is a natural assumption when the ordering of vertices does not matter. There is a deep structural theory for exchangeable undirected graphs, which extends to the directed case via measurable objects known as digraphons. Using digraphons, we first show how to construct models for exchangeable directed graphs, including special cases such as tournaments, linear orderings, directed acyclic graphs, and partial orderings. We then show how to construct priors on digraphons via the infinite relational digraphon model (di-IRM), a new Bayesian nonparametric block model for exchangeable directed graphs, and demonstrate inference on synthetic data.