Reinforcement learning (RL) is a general framework for modeling sequential decision making problems, at the core of which lies the dilemma of exploitation and exploration. An agent failing to explore systematically will inevitably fail to learn efficiently. Optimism in the face of uncertainty (OFU) is a conventionally successful strategy for efficient exploration. An agent following the OFU principle explores actively and efficiently. However, when applied to model-based RL, it involves specifying a confidence set of the underlying model and solving a series of nonlinear constrained optimization, which can be computationally intractable. This paper proposes an algorithm, Bayesian optimistic optimization (BOO), which adopts a dynamic weighting technique for enforcing the constraint rather than explicitly solving a constrained optimization problem. BOO is a general algorithm proved to be sample-efficient for models in a finite-dimensional reproducing kernel Hilbert space. We also develop techniques for effective optimization and show through some simulation experiments that BOO is competitive with the existing algorithms.

In this paper, we consider the RL in Markov decision processes (MDPs), where the agent observes the state of the environment at each timestep and makes decisions accordingly.

Reinforcement learning (RL) is a general framework for modeling sequential decision making problems, at the core of which lies the dilemma of exploitation and exploration. An agent failing to explore systematically will inevitably fail to learn efficiently. Optimism in the face of uncertainty (OFU) is a conventionally successful strategy for efficient exploration. An agent following the OFU principle explores actively and efficiently. However, when applied to model-based RL, it involves specifying a confidence set of the underlying model and solving a series of nonlinear constrained optimization, which can be computationally intractable.

We point out important problems with the common practice of using the best single model performance for comparing deep learning architectures, and we propose a method that corrects these flaws. Each time a model is trained, one gets a different result due to random factors in the training process, which include random parameter initialization and random data shuffling. Reporting the best single model performance does not appropriately address this stochasticity. We propose a normalized expected best-out-of-$n$ performance ($\text{Boo}_n$) as a way to correct these problems.

As the emergence of AI grows stronger, many have begun to conflate this technology with the previous wave of automation that has gone before it. Chris Boos, CEO and founder of Arago refutes this stance, and explains how publishers and other businesses are already practically applying artificial intelligence to their brands. Chris Boos is featured prominently on the homepage of the Arago website, along with an immediate introduction to his own personal mission statement: 'empowering human potential, freeing up time for creativity and innovative thinking through artificial intelligence (AI).' It's a notably human-focussed shop window for the now 23 year old German company, which specialises in helping clients to leverage artificial intelligence to reinvent their business models for the digital age. One anecdote Boos tells us is around the computer game Civilisation, where the turn-based strategy has been used to facilitate greater machine learning within an investment fund. As he talks about the ability of games to educate both children and adults, it's clear that Boos is taking a'ground up' approach to building digital intelligence.

Warnings about artificial intelligence launching World War III--including a few flares sent up by Elon Musk--are an unfair scourge on an AI sector that sees itself making life easier and helping traditional companies survive. That's the view of Chris Boos, chief executive officer of Germany-based software firm Arago, who told MarketWatch in an interview that anything produced by a process can, should and will be run by AI, allowing human beings to be the creative thinkers and doers they were designed to be. Arago advises mostly non-tech, established-economy Fortune 500 businesses on their AI adoption. "Within the next 2-3 years AI will be able to run any business process, which makes AI one -- potentially the only one -- defensive measure the established economy has against intrusion from the high-tech world," said Boos. For now, the sci-fi hyperbole can wait.

SAP's Christian Boos had a light-bulb moment this year at SAPPHIRE NOW. The global business development expert for machine learning was due to hold a 20-minute meeting with a long-time SAP customer, but 20 minutes soon turned into 90. "They just kept on reeling off the use cases for machine learning at their company," Boos says. This and many other conversations at the event fed his conviction that, far from being "just another bandwagon," machine learning is a topic that will help decide many companies' future. While the consumer market is already brimming with highly advanced artificial intelligence (AI) and machine learning products, many enterprises are only just starting to embrace these technologies.

If an AI could rule a world, would you trust it to manage your IT systems? German software company Arago is hoping you will. The developer of IT automation system Hiro (short for Human Intelligence Robotically Optimized) has been teaching its software how to play Freeciv, an open source computer strategy game inspired by Sid Meier's Civilization series of games, and in the process is learning to make IT management more fun. Hiro is an AI-based automation system that usually sits on top of other IT service management tools. Unlike script-based systems, it learns from its users how best to manage a company's IT systems.

In this paper, we translate a model of value-based emo- tional agents into an architecture for narrative characters and we validate it in a narrative scenario. The advantage of using such model is that different moral behaviors can be obtained as a consequence of the emotional ap- praisal of moral values, a desirable feature for digital storytelling techniques.