Fixed-time strategy is a common approach in signal traffic control, characterized by simple and periodic signal plans that are easy to implement without detection mechanisms. A major step in the design of such plans refers to the grouping of the day hours such that the same plan applies for several consecutive hours. The efficacy of the plan, measured by vehicle delays, relies on the matching of traffic with the fixed plan. Accordingly, the time-of-day breakpoints between plans are selected based on the variability of the traffic within each group. The paper studies the selection of time-of-day breakpoints (TODs) based on real traffic characteristics from two cities. Motivated by the Google Green Light project, this study presents an approach to compute TODs based on aggregated traffic statistics computed from anonymized trajectories from navigation applications. We evaluate an optimal dynamic programming algorithm to compute time-of-day breakpoints at an intersection, based on traffic variability among hours. We analyze typical forms of efficient time-of-day breakpoints and examine the impact of the number of daily plans on the ability to predict traffic behavior. We refer to various metrics to measure the variability of the traffic within sets of hours concerning the amount of traffic and its distribution among various movements. We measure the dissimilarity of the time-of-day breakpoints when computed for the different metrics. We also address the joint computation of TODs in adjacent intersections to improve coordination potential.

Optimization of machine learning models is commonly performed through stochastic gradient updates on randomly ordered training examples. This practice means that each fraction of an epoch comprises an independent random sample of the training data that may not preserve informative structure present in the full data. We hypothesize that the training can be more effective with self-similar arrangements that potentially allow each epoch to provide benefits of multiple ones. We study this for “matrix factorization” – the common task of learning metric embeddings of entities such as queries, videos, or words from example pairwise associations. We construct arrangements that preserve the weighted Jaccard similarities of rows and columns and experimentally observe training acceleration of 3%-37% on synthetic and recommendation datasets. Principled arrangements of training examples emerge as a novel and potentially powerful enhancement to SGD that merits further exploration.

Distances in a network capture relations between nodes and are the basis of centrality, similarity, and influence measures. Often, however, the relevance of a node u to a node v is more precisely measured not by the magnitude of the distance, but by the number of nodes that are closer to v than u. That is, by the rank of u in an ordering of nodes by increasing distance from v. We identify and address fundamental challenges in rank-based graph mining. We first consider single-source computation of reverse-ranks and design a "Dijkstra-like" algorithm which computes nodes in order of increasing approximate reverse rank while only traversing edges adjacent to returned nodes. We then define reverse-rank influence, which naturally extends reverse nearest neighbors influence [Korn and Muthukrishnan 2000] and builds on a well studied distance-based influence. We present near-linear algorithms for greedy approximate reverse-rank influence maximization. The design relies on our single-source algorithm. Our algorithms utilize near-linear preprocessing of the network to compute all-distance sketches. As a contribution of independent interest, we present a novel algorithm for computing these sketches, which have many other applications, on multi-core architectures.