In preference-based reinforcement learning (RL), an agent interacts with the environment while receiving preferences instead of absolute feedback. While there is increasing research activity in preference-based RL, the design of formal frameworks that admit tractable theoretical analysis remains an open challenge. Building upon ideas from preference-based bandit learning and posterior sampling in RL, we present Dueling Posterior Sampling (DPS), which employs preference-based posterior sampling to learn both the system dynamics and the underlying utility function that governs the user's preferences. Because preference feedback is provided on trajectories rather than individual state/action pairs, we develop a Bayesian approach to solving the credit assignment problem, translating user preferences to a posterior distribution over state/action reward models. We prove an asymptotic no-regret rate for DPS with a Bayesian logistic regression credit assignment model; to our knowledge, this is the first regret guarantee for preference-based RL. We also discuss possible avenues for extending this proof methodology to analyze other credit assignment models. Finally, we evaluate the approach empirically, showing competitive performance against existing baselines.

Interference between pharmacological substances can cause serious medical injuries. Correctly predicting so-called drug-drug interactions (DDI) does not only reduce these cases but can also result in a reduction of drug development cost. Presently, most drug-related knowledge is the result of clinical evaluations and post-marketing surveillance; resulting in a limited amount of information. Existing data-driven prediction approaches for DDIs typically rely on a single source of information, while using information from multiple sources would help improve predictions. Machine learning (ML) techniques are used, but the techniques are often unable to deal with skewness in the data. Hence, we propose a new ML approach for predicting DDIs based on multiple data sources. For this task, we use 12,000 drug features from DrugBank, PharmGKB, and KEGG drugs, which are integrated using Knowledge Graphs (KGs). To train our prediction model, we first embed the nodes in the graph using various embedding approaches. We found that the best performing combination was a ComplEx embedding method creating using PyTorch-BigGraph (PBG) with a Convolutional-LSTM network and classic machine learning-based prediction models. The model averaging ensemble method of three best classifiers yields up to 0.94, 0.92, 0.80 for AUPR, F1-score, and MCC, respectively during 5-fold cross-validation tests.

The grocery story business is beginning to use AI to try to gain a competitive edge. Salt Lake City-based Associated Food Stores (AFS), for example, has 500 stores in the western and southwestern US. It found itself dealing with a growing number of SKUs that stores managers were having difficulty tracking and prioritizing, according to an account in ChainStoreAge. AFS began using an AI solution from CB4 to analyze point of sale data, to identify when physical issues in a store are hold back sales. These could be products not easily visible and out of stock conditions.

As the Veterans Affairs Department's inaugural Director of Artificial Intelligence, Gil Alterovitz aims to leverage the emerging technology and the agency's cornucopia of data to proactively anticipate and tackle problems afflicting veterans like never before. In a conversation with Nextgov, Alterovitz detailed his present efforts and future-facing vision to support VA in executing that mission. "Nowhere in the country is there such potential for research to be developed and translated into clinical care so quickly. In this case, it's to help our special population of veterans … and those patients have actually asked us to deal with their needs," Alterovitz said. "We really want to be the go-to place for veterans through AI research and development--so instead of reacting, we can really anticipate their needs."

You can now remove your local bar from the list of places where you wouldn't expect to see facial recognition tech. DataSparQ, a data science company based in Britain, has developed a system that keeps track of people's places in line at the bar. AI bar uses cameras and artificial intelligence to recognize when a customer approaches the bar to order, capturing their face and logging it into a running queue. This queue is displayed on a screen mounted at the bar, along with a live feed that circles each person's face. The system can also flag faces that it thinks might be under 25, prompting the bartender to check their ID.

In addition, we'll inform you about our many upcoming events in Boston, NYC, San Francisco, and London. And keep a lookout for special discount codes, only available to our newsletter subscribers! We're Proud to Have Their Best and Brightest in Attendance

You could be forgiven for thinking that AI will soon replace human physicians based on headlines such as "The AI Doctor Will See You Now," "Your Future Doctor May Not Be Human," and "This AI Just Beat Human Doctors on a Clinical Exam." But experts say the reality is more of a collaboration than an ousting: Patients could soon find their lives partly in the hands of AI services working alongside human clinicians. There is no shortage of optimism about AI in the medical community. But many also caution the hype surrounding AI has yet to be realized in real clinical settings. There are also different visions for how AI services could make the biggest impact.

A collaborative research team claims their artificially intelligent system should be recognized as the rightful inventor of two innovative designs, in a potentially disruptive development in patent law. Patent law is complicated even at the best of times, but a new project led by researchers from the University of Surrey could make it more convoluted still. Called the Artificial Inventor Project, the initiative is "seeking intellectual property rights for the autonomous output of artificial intelligence." As BBC reports, the researchers are claiming that an artificially intelligent system named DABUS is the rightful inventor of two designs, namely a complex, fractal-like system of interlocking food containers and a rhythmic warning light for attracting extra attention. To that end, the researchers are filing patents on behalf of DABUS with the respective patent bodies in the United States, the United Kingdom, and the European Union.

At the final session of the 2019 Space Symposium in Colorado Springs, attendees straggled into a giant ballroom to listen to an Air Force official and a National Geospatial-Intelligence Agency (NGA) executive discuss, as the panel title put it, "Enterprise Disruption." The presentation stayed as vague as the title until a direct question from the audience seemed to make the panelists squirm. Just how good, the person wondered, had the military and intelligence communities' algorithms gotten at interpreting data and taking action based on that analysis? They pointed out that the commercial satellite industry has software that can tally shipping containers on cargo ships and cars in parking lots soon after their pictures are snapped in space. "When will the Department of Defense have real-time, automated, global order of battle?" they asked. "That's a great question," said Chirag Parikh, director of the NGA's Office of Sciences and Methodologies.

How can we prevent the dream of digitalisation for people's good from turning into a threat or pure unrealistic futurology? Can pervasive technology become invasive? Could AI turn against patients and health professionals? The more digital technologies accompany patients and physicians, the more opportunities for better outcomes, but also concerns, especially those related to data security. China's social credit system raises many ethical questions, some chatbots available online are misleading, data gathered in social media can be used to influence political decisions, fake news distributed on the internet threaten human's health and lives.