Computation of Dynamic Equilibria in Series-Parallel Networks

We consider dynamic equilibria for flows over time under the fluid queuing model. In this model, queues on the links of a network take care of flow propagation. Flow enters the network at a single source and leaves at a single sink. In a dynamic equilibrium, every infinitesimally small flow particle reaches the sink as early as possible given the pattern of the rest of the flow. Although this model has been examined for many decades, progress has been relatively recent. In particular, the derivatives of dynamic equilibria have been characterized as thin flows with resetting, which allows for more structural results. Our two main results are based on the formulation of thin flows with resetting as a linear complementarity problem and its analysis. We present a constructive proof of existence for dynamic equilibria if the inflow rate is right-monotone. The complexity of computing thin flows with resetting, which occurs as a subproblem in this method, is still open. We settle it for the class of two-terminal, series-parallel networks by giving a recursive algorithm that solves the problem for all flow values simultaneously in polynomial time.

Parallel systems for social and spatial reasoning within the cortical apex

What is the cognitive and neural architecture of systems for high-level reasoning and memory in humans? We ask this question using deep neuroimaging of individual human brains on various tasks involving reasoning and memory about familiar people, places, and objects. We find that thinking about people and places elicits responses in distinct but parallel networks within high-level association cortex, spanning the frontal, parietal, and temporal lobes. Person- and place-preferring brain regions were systematically yoked across multiple cortical zones. These regions were strongly category-selective, across visual, semantic, and episodic tasks, and were specifically functionally connected to other parts of association cortex with similar category preferences. These results demonstrate that selectivity for content domain is a widespread feature of high-level association cortex in humans. They support a theoretical model in which reasoning about people and places are supported by parallel cognitive and neural mechanisms.

On the Optimal Control of Parallel Processing Networks with Resource Collaboration and Multitasking

We study scheduling control of parallel processing networks in which some resources need to simultaneously collaborate to perform some activities and some resources multitask. Resource collaboration and multitasking give rise to synchronization constraints in resource scheduling when the resources are not divisible, that is, when the resources cannot be split. The synchronization constraints affect the system performance significantly. For example, because of those constraints, the system capacity can be strictly less than the capacity of the bottleneck resource. Furthermore, the resource scheduling decisions are not trivial under those constraints. For example, not all static prioritization policies retain the maximum system capacity, and the ones that retain the maximum system capacity do not necessarily minimize the delay (or, in general, the holding cost). We study optimal scheduling control of a class of parallel networks and propose a dynamic prioritization policy that retains the maximum system capacity and is asymptotically optimal in diffusion scale and a conventional heavy-traffic regime with respect to the expected discounted total holding cost objective.

Reciprocal repulsions instruct the precise assembly of parallel hippocampal networks

Mammalian medial and lateral hippocampal networks preferentially process spatial- and object-related information, respectively. However, the mechanisms underlying the assembly of such parallel networks during development remain largely unknown. Our study shows that, in mice, complementary expression of cell surface molecules teneurin-3 (Ten3) and latrophilin-2 (Lphn2) in the medial and lateral hippocampal networks, respectively, guides the precise assembly of CA1-to-subiculum connections in both networks. In the medial network, Ten3-expressing (Ten3+) CA1 axons are repelled by target-derived Lphn2, revealing that Lphn2- and Ten3-mediated heterophilic repulsion and Ten3-mediated homophilic attraction cooperate to control precise target selection of CA1 axons. In the lateral network, Lphn2-expressing (Lphn2+) CA1 axons are confined to Lphn2+ targets via repulsion from Ten3+ targets. Our findings demonstrate that assembly of parallel hippocampal networks follows a “Ten3→Ten3, Lphn2→Lphn2” rule instructed by reciprocal repulsions.

Vps34 and TOR Kinases Coordinate HAC1 mRNA Translation in the presence or absence of Ire1-dependent Splicing

In the budding yeast Saccharomyces cerevisiae an mRNA, called HAC1, exists in a translationally repressed form in the cytoplasm. Under conditions of cellular stress, such as when unfolded proteins accumulate inside the endoplasmic reticulum (ER), an RNase Ire1 removes an intervening sequence (intron) from the HAC1 mRNA by non-conventional cytosolic splicing. Removal of the intron results in translational de-repression of HAC1 mRNA and production of a transcription factor that activates expressions of many enzymes and chaperones to increase the protein-folding capacity of the cell. Here, we show that Ire1-mediated RNA cleavage requires Watson-Crick base pairs in two RNA hairpins, which are located at the HAC1 mRNA exon-intron junctions. Then, we show that the translational de-repression of HAC1 mRNA can occur independent of cytosolic splicing. These results are obtained from HAC1 variants that translated an active Hac1 protein from the un-spliced mRNA. Additionally, we show that the phosphatidylinositol-3-kinase Vps34 and the nutrient-sensing kinases TOR and GCN2 are key regulators of HAC1 mRNA translation and consequently the ER stress responses. Collectively, our data suggest that the cytosolic splicing and the translational de-repression of HAC1 mRNA are coordinated by unique and parallel networks of signaling pathways.

A Classification Method for the Cellular Images Based on Active Learning and Cross-Modal Transfer Learning

In computer-aided diagnosis (CAD) systems, the automatic classification of the different types of the human epithelial type 2 (HEp-2) cells represents one of the critical steps in the diagnosis procedure of autoimmune diseases. Most of the methods prefer to tackle this task using the supervised learning paradigm. However, the necessity of having thousands of manually annotated examples constitutes a serious concern for the state-of-the-art HEp-2 cells classification methods. We present in this work a method that uses active learning in order to minimize the necessity of annotating the majority of the examples in the dataset. For this purpose, we use cross-modal transfer learning coupled with parallel deep residual networks. First, the parallel networks, which take simultaneously different wavelet coefficients as inputs, are trained in a fully supervised way by using a very small and already annotated dataset. Then, the trained networks are utilized on the targeted dataset, which is quite larger compared to the first one, using active learning techniques in order to only select the images that really need to be annotated among all the examples. The obtained results show that active learning, when mixed with an efficient transfer learning technique, can allow one to achieve a quite pleasant discrimination performance with only a few annotated examples in hands. This will help in building CAD systems by simplifying the burdensome task of labeling images while maintaining a similar performance with the state-of-the-art methods.

On the zagreb connection indices of hex and honeycomb networks

Regular plane tessellations can easily be constructed by repeating regular polygons. This design is of extreme importance for direct interconnection networks as it yields high overall performance. The honeycomb and the hexagonal networks are two such popular mesh-derived parallel networks. The first and second Zagreb indices are among the most studied topological indices. We now consider analogous graph invariants, based on the second degrees of vertices, called Zagreb connection indices. The main objective of this paper is to compute these connection indices for the Hex, Hex derived and some Honeycomb networks.