Social media generated by location-services-enabled cellular devices produce enormous amounts of location-based content. Spatiotemporal analysis of such data facilitate new ways of modeling human behavior and mobility patterns. In this paper, we use over 10 millions geo-tagged tweets from the city of Los Angeles as observations of human movement and apply them to understand the relationships of geographical regions, neighborhoods and gang territories. Using a graph based-representation of street gang territories as vertices and interactions between them as edges, we train a machine learning classifier to tell apart rival and non-rival links. We correctly identify 89% of the true rivalry network, which beats a standard baseline by about 30%. Looking at larger neighborhoods, we were able to show that the direction of displacement, i.e, the distribution of movement direction, can be used as a profile to identify physical (or geographic) barriers when it is not uniform. Finally, considering the temporal dimension of tweets, we detect events taking place around the city by identifying irregularities in tweeting patterns.

A candidate framework for integration of theoretical, modeling and experimental approaches to understanding emotional intelligence is described. The framework includes three elements of a new kind that enable representation of emotional cognition: an emotional state, an appraisal, and a moral schema. These elements are integrated with the weak semantic cognitive map representing the values of emotional appraisals. The framework is tested on interpretation of results obtained in two new experimental paradigms that reveal general features of human emotional cognition, such as the emergence of subjectively perceived persistent roles of individual virtual actors. Implications concern heterogeneous human-robot teams.

The Cognitive Calculus is the result of decades of work in artificial intelligence, psychology, linguistics, and systems engineering. It is a notation we use to model human cogni-tion and to guide our development of a Cognitive Database (CDB), our evolving computer model of human cognition. The Cognitive Calculus acknowledges human memory (labeled as Total Memory, TM) as the central component of human intelligence. TM is composed of all memory subsystems including: short term memory (STM – our working memory), episodic memory (EM – our life history), and abstracted memory (AM – our mental models). The Cognitive Calculus defines the basic element of TM as a memory node (MemNo) along with a set of operations intrinsic to the creation, retrieval, update, and deletion of MemNos. All content in all of TM is a sensory input, an effector activity, a cognitive activity, an abstracted set (T) or an abstracted sequence (Q). This paper describes how a human cognitive system takes the experiences of life in as inputs to STM, passes those inputs through STM into EM, finds patterns in EM for creating abstractions in AM, and then uses those abstractions of the past to comprehend its current experience. The paper ends with guidance concerning the things that must be in a Cognitive Database so that a computer can better model human cognition.

Recently, an increasing number of mobile users are eagerly using the cellular network in data applications. In particular, multimedia downloads generated by Internet-capable smart phones and other portable devices has been widely recognized as the major source for strains in cellular networks, to a degree where service quality for all users is significantly impacted. In this paper we explore the novel concept of proactive content caching using evolutionary algorithms inspired by the inherent predictability of the mobile user behavior. Users can then use the cached version of the content in order to achieve a better user experience and reduce the peak-to-average ratio in mobile networks, especially during peak hours of the day. Finally, we confirm the merits of the proposed scheduler using real data traces of different user's requests and Wi-Fi availability. The results after applying the proposed scheduling algorithm show that up to 70% of the user content requests can be fulfilled i.e. the content were successfully cached before request. We also observe that proposed scheduler outperforms a baseline scheduler based on simulated annealing.

This short position paper claims that computational neuroscience should refocus on the study of multiscale spatiotemporal shapes (STS) of activity in large neural populations. Instead of naive engineering metaphors, which view the brain as a signal-processing channel traversed by "information", or neo-Behaviorist probabilistic frameworks, where it is a "gray box" tuned by environmental distributions, new theories should resolutely promote mechanistic, complex systems models. In this paradigm, massively recurrent networks should support the spontaneous (and triggered) emergence of intrinsic dynamical regimes, made of myriads of correlated electrophysiological signals—not unlike other collective biological phenomena such as bird flocks, insect constructions, or morphogenesis. "Neuron flocking", for its part, must happen in phase space and across a complex network topology: Can we characterize the "shapes" and composition laws of these mind states, upon which high-level symbolic computing can ultimately rest?

Subfields of artificial intelligence often diversify from a core idea. For example, deep learning networks, models in computational neuroscience, and neuroevolution all take inspiration from biological neural networks as a potential pathway to AI. Most researchers choose to pursue the subfield (and by extension, abstraction) they see as most promising for leading to AI, which naturally results in significant debate and disagreement among researchers as to what abstraction is best. A better understanding and less polarized debate may result from a clear presentation and discussion of abstractions by their most knowledgeable proponents. These insights motivated bringing together researchers from fields that abstract AI at different levels or in different ways to disperse knowledge, and to critically examining the value and promise of different abstractions. Thus this AAAI symposium, How Intelligence Should be Abstracted in AI, consisted of a diverse and multidisciplinary group of AI researchers interested in discussing and comparing different abstractions of both intelligence and processes that might create it.

The emphasis of multilevel modeling techniques in the neurosciences has led to an increased need for large-scale, computationally-accessible databases containing neuroscientific data. Despite this, such databases are not being populated at a rate commensurate with their demand amongst Neuroinformaticians. The reasons for this are common to scientific database curation in general, namely, limitation of resources. Much of neuroscience's long tradition of research has been documented in computationally inaccessible formats, such as the \emph{pdf}, making large-scale data extraction laborious and expensive. Here, we present a system for alleviating one bottleneck in the workflow for curating a typical knowledge base of neuroscience-related information. Finna is designed to rank-order the composite paragraphs of a publication that is predicted to contain information relevant to a knowledge base, in terms of the probability that each documents relevant data. We were able to achieve excellent performance with our classifier (AUC > 0.90) on our manually-curated neuroscience document corpus. Our approach would allow curators to read only a median of 2 paragraphs for each document, in order to identify information relevant to a neuron-related knowledge base. To our knowledge, this is the first system of its kind, and will be a useful baseline for developing similar resources for the neurosciences, and curation in general.

With recent advances in high throughput technology, researchers often find themselves running a large number of hypothesis tests (thousands+) and esti- mating a large number of effect-sizes. Generally there is particular interest in those effects estimated to be most extreme. Unfortunately naive estimates of these effect-sizes (even after potentially accounting for multiplicity in a testing procedure) can be severely biased. In this manuscript we explore this bias from a frequentist perspective: we give a formal definition, and show that an oracle estimator using this bias dominates the naive maximum likelihood estimate. We give a resampling estimator to approximate this oracle, and show that it works well on simulated data. We also connect this to ideas in empirical Bayes.

Belief Propagation has been widely used for marginal inference, however it is slow on problems with large-domain variables and high-order factors. Previous work provides useful approximations to facilitate inference on such models, but lacks important anytime properties such as: 1) providing accurate and consistent marginals when stopped early, 2) improving the approximation when run longer, and 3) converging to the fixed point of BP. To this end, we propose a message passing algorithm that works on sparse (partially instantiated) domains, and converges to consistent marginals using dynamic message scheduling. The algorithm grows the sparse domains incrementally, selecting the next value to add using prioritization schemes based on the gradients of the marginal inference objective. Our experiments demonstrate local anytime consistency and fast convergence, providing significant speedups over BP to obtain low-error marginals: up to 25 times on grid models, and up to 6 times on a real-world natural language processing task.

The seemingly stochastic transient dynamics of neocortical circuits observed in vivo have been hypothesized to represent a signature of ongoing stochastic inference. In vitro neurons, on the other hand, exhibit a highly deterministic response to various types of stimulation. We show that an ensemble of deterministic leaky integrate-and-fire neurons embedded in a spiking noisy environment can attain the correct firing statistics in order to sample from a well-defined target distribution. We provide an analytical derivation of the activation function on the single cell level; for recurrent networks, we examine convergence towards stationarity in computer simulations and demonstrate sample-based Bayesian inference in a mixed graphical model. This establishes a rigorous link between deterministic neuron models and functional stochastic dynamics on the network level.