History doesn't exactly repeat itself, but it does run in cycles. One of the most robust theories of such cycles was articulated by economic historian Carlota Perez, in her influential book Technological Revolutions and Financial Capital: The Dynamics of Bubbles and Golden Ages (Edward Elgar, 2002). It suggests that humanity can get through the current period of upheaval and economic malaise and enter a new "golden age" of broad economic growth, if the world's key decision makers act in concert to help foster one. This may seem far-fetched, but it's happened four times before. We are in the midst of the fifth great surge (as Perez calls them) of technological and economic change since the Industrial Revolution. The last one, the age of oil, automobiles, and mass production, lasted most of the 20th century and still shapes many people's attitudes. Our current surge started around 1970 and has rolled out information and communications technology around the world: It is the age of the computer and the Internet (see Exhibit 1). Each of these surges follows the same broad pattern. First, there is a wave of major new technologies, leading to dramatic changes in industrial production and daily life. For about 20 to 30 years, in a period that Perez calls installation, these technologies are funded largely by speculative investment chasing rapid returns. This age of widening wealth disparity leads to a bubble, which bursts in spectacular fashion, and is followed by a crisis period that Perez calls the turning point. This phase of economic and social turbulence has varied in length from two years to 17. Many efforts to get back to normal are made, usually involving the regulation of financial excesses or the stimulation of production and employment.

Rolls-Royce and Google on the same boat with navigating a path towards an AI driven autonomous future. Google is shipping out its Cloud Machine Learning Engine to Rolls-Royce to help make autonomous ships a reality. The hope is that the Google Cloud Machine Learning Engine will be able to train Rolls-Royce's intelligent awareness systems, an AI-based classification system for detecting, identifying, and tracking the objects that a vessel could come across at sea. Signed at the Google Cloud Summit in Sweden, the agreement will help the company to create bespoke machine learning models that'll be able to uncover valuable insights from data sets that are created by Rolls-Royce. Karno Tenovuo, Rolls-Royce, SVP Ship Intelligence said: "While intelligent awareness systems will help to facilitate an autonomous future, they can benefit maritime businesses right now making vessels and their crews safer and more efficient. By working with Google Cloud we can make these systems better faster, saving lives."

If there's anyone who can talk about the rise of fintech (financial technology) and where the sector is going, its Microsoft's director of financial services, Richard Peers. He's been working for the legacy tech company for the past 21 years, and it's given him some advantages. And one of the advantages of being around for that amount of time is as the industry changes, you can change your roles," he told Verdict. Six years ago he was working with Microsoft's teams that were working with the banks developing mobile apps. In doing that, I met the whole wave of the need to modernise banks, the need to improve the customer experience, deal with the frustration employees had with the systems, and what was being done to improve the operational capabilities. Ahead of his speech at Lendit Europe 2017 next week, Peers (left) spoke to Verdict about the rise and future of fintech, why banks need to learn from Chinese tech companies and his thoughts on artificial intelligence (AI). "After the banking crisis, you had a lot of computer science graduates who couldn't get jobs and a lot of bankers who really understood banking were being made redundant or choosing to leave.

Development of black-box modeling techniques, like support vector machine (SVM), neural networks, etc., has shown rather rapid in the past decades (Yuan et al., 2016; Zhao et al., 2015; Wu et al., 2013). This sort of techniques, compared to white-box modeling methods (also called mechanism-based modeling or first-principles modeling), works without any need of knowing the internal structure or details on variables interaction in systems considered, so they are suited to describe extremely complex objectives, such as human brain (Khosrowabadi et al., 2014), black hole (Grumiller et al., 2012), integrated industrial processes (Gao et al., 2012) and so on. Essentially, blackbox modeling is an input-output data-based approach, and the model precision mainly depends on data quality, model structure and parameters identification algorithm. In order to develop high-precision black-box models, it always needs reliable and representative data, smart mathematical treatment and efficient identification algorithms. All of these are challenging the development of the black-box modeling techniques.

In a recent note [8], the author provides a counterexample to the global convergence of what his work refers to as "the DSOS and SDSOS hierarchies" for polynomial optimization problems (POPs) and purports that this refutes claims in our extended abstract [4] and slides in [3]. The goal of this paper is to clarify that neither [4], nor [3], and certainly not our full paper [5], ever defined DSOS or SDSOS hierarchies as it is done in [8]. It goes without saying that no claims about convergence properties of the hierarchies in [8] were ever made as a consequence. What was stated in [4,3] was completely different: we stated that there exist hierarchies based on DSOS and SDSOS optimization that converge. This is indeed true as we discuss in this response. We also emphasize that we were well aware that some (S)DSOS hierarchies do not converge even if their natural SOS counterparts do. This is readily implied by an example in our prior work [5], which makes the counterexample in [8] superfluous. Finally, we provide concrete counterarguments to claims made in [8] that aim to challenge the scalability improvements obtained by DSOS and SDSOS optimization as compared to sum of squares (SOS) optimization. [3] A. A. Ahmadi and A. Majumdar. DSOS and SDSOS: More tractable alternatives to SOS. Slides at the meeting on Geometry and Algebra of Linear Matrix Inequalities, CIRM, Marseille, 2013. [4] A. A. Ahmadi and A. Majumdar. DSOS and SDSOS optimization: LP and SOCP-based alternatives to sum of squares optimization. In proceedings of the 48th annual IEEE Conference on Information Sciences and Systems, 2014. [5] A. A. Ahmadi and A. Majumdar. DSOS and SDSOS optimization: more tractable alternatives to sum of squares and semidefinite optimization. arXiv:1706.02586, 2017. [8] C. Josz. Counterexample to global convergence of DSOS and SDSOS hierarchies. arXiv:1707.02964, 2017.

We consider the problem of classifying business process instances based on structural features derived from event logs. The main motivation is to provide machine learning based techniques with quick response times for interactive computer assisted root cause analysis. In particular, we create structural features from process mining such as activity and transition occurrence counts, and ordering of activities to be evaluated as potential features for classification. We show that adding such structural features increases the amount of information thus potentially increasing classification accuracy. However, there is an inherent trade-off as using too many features leads to too long run-times for machine learning classification models. One way to improve the machine learning algorithms' run-time is to only select a small number of features by a feature selection algorithm. However, the run-time required by the feature selection algorithm must also be taken into account. Also, the classification accuracy should not suffer too much from the feature selection. The main contributions of this paper are as follows: First, we propose and compare six different feature selection algorithms by means of an experimental setup comparing their classification accuracy and achievable response times. Second, we discuss the potential use of feature selection results for computer assisted root cause analysis as well as the properties of different types of structural features in the context of feature selection.

Oracle OpenWorld -- Oracle today announced new artificial intelligence-based apps for finance, human resources, supply chain, manufacturing, commerce, customer service, marketing, and sales professionals. The new Oracle Adaptive Intelligent Apps are built into the existing Oracle Cloud Applications to deliver the industry's most powerful AI-based modern business applications. "The new Adaptive Intelligent Apps enable business users from across the organizations to quickly and easily take advantage of the latest advancements in artificial intelligence," said Steve Miranda, executive vice president of applications development, Oracle. "To make this possible we have eliminated the need for more integrations and embedded AI capabilities across Oracle Cloud Applications. The new AI capabilities combine first- and third-party data with advanced machine learning and sophisticated decision science to deliver the industry's most powerful AI-based modern business applications." The new Oracle Adaptive Intelligent Apps deliver immediate impact within Oracle Enterprise Resource Planning Cloud, Oracle Human Capital Management Cloud, Oracle Supply Chain Management Cloud, and the Oracle Customer Experience Cloud by providing smart and timely insights to end users.

The Internet of Things is already part of our history. Surpassed, at least in the way it is still described in many conferences and papers. New technology advancements and human-machine collaboration models are redefying its potential, centered around cloud and artificial intelligence. Let's see some factors which are shaping this new kind of "Internet of"… which is deeply transforing the way Humans and Things interact and collaborate! At Microsoft we imagine a new world where conversation and vision is the new computer interface, able to proactively and personally serve our needs.

Cylance polled 652 IT decision makers in the U.S., UK, Germany and France, and found that optimism about the value of artificial intelligence-powered solutions in the enterprise is high and plans to continue investment in the technology are widespread. AI is already making a significant impact in the enterprise, analyzing trends in security, operational efficiency, marketing, employee perceptions, and other areas. Organizations are already investing in AI, and this will only increase: Nearly all of the IT decision makers surveyed said they are either currently spending on AI-powered solutions or planning to invest in them in the next two years; 60 percent already have AI in place. Additionally, 79 percent say AI is a top priority for their boards and C-suite executives. For security teams, AI is moving the needle: Seventy-seven percent have prevented more breaches following their use of AI-powered tools and 81 percent say AI was detecting threats before their security teams could.

Historical data – it's a bad system, but it's the best we've got, right? It's biased, out of date, and based on the flawed assumption that the future will look like the past. Historical data is far from ideal for training your artificial intelligence (AI) systems on. We've collated some expert advice to answer those questions for you here: "If you really want an AI system that delivers good business value, it's got to be forward looking, and therefore it must look at real time execution data." "If it's not then it's always looking in the rear-view mirror, which doesn't help me make good decisions which increase my revenue and decrease my cost."