Search

Driven by real-life applications in geo-social networks, we study the problem of computing radius-bounded k-cores (RB-k-cores) that aims to find communities satisfying both social and spatial constraints. In particular, the model k-core (i.e., the subgraph where each vertex has at least k neighbors) is used to ensure the social cohesiveness, and a radius-bounded circle is used to restrict the locations of users in an RB-k-core. We explore several algorithmic paradigms to compute RB-k-cores, including a triple-vertex-based paradigm, a binary-vertex-based paradigm, and a paradigm utilizing the concept of rotating circles. The rotating-circle-based paradigm is further enhanced by several pruning techniques to achieve better efficiency. In addition, to find representative RB-k-cores, we study the diversified radius-bounded k-core search problem, which finds t RB-k-cores to cover the most number of vertices. We first propose a baseline algorithm that identifies the distinctive RB-k-cores after finding all the RB-k-cores. Beyond this, we design algorithms that can efficiently maintain the top-t candidate RB-k-cores and also achieve a guaranteed approximation ratio. Experimental studies on both real and synthetic datasets demonstrate that our proposed techniques can efficiently compute (diversified) RB-k-cores. Moreover, our techniques can be used to compute the minimum-circle-bounded k-core and significantly outperform the existing techniques.

We analyze a tree search problem with an underlying Markov decision process, in which the goal is to identify the best action at the root that achieves the highest cumulative reward. We present a new tree policy that optimally allocates a limited computing budget to maximize a lower bound on the probability of correctly selecting the best action at each node. Compared to widely used Upper Confidence Bound (UCB) tree policies, the new tree policy presents a more balanced approach to manage the exploration and exploitation trade-off when the sampling budget is limited. Furthermore, UCB assumes that the support of reward distribution is known, whereas our algorithm relaxes this assumption. Numerical experiments demonstrate the efficiency of our algorithm in selecting the best action at the root., Comment: Submitted to IEEE Transactions on Automatic Control

This paper studies a search problem in which a consumer is initially aware of only a few products. At every point in time, the consumer then decides between searching among alternatives of which he is already aware and discovering more products. I show that the optimal policy for this search and discovery problem is fully characterized by tractable reservation values. Moreover, I prove that a predetermined index fully specifies the purchase decision of a consumer following the optimal search policy. Finally, a comparison highlights differences to classical random and directed search. This paper was accepted by Dmitri Kuksov, marketing.

This article focuses on the target search problem in a curve-shape area using multiple unmanned aerial vehicles (UAVs), with the demand for obtaining the maximum cumulative detection reward, as well as the constraint of maneuverability and obstacle avoidance. First, the prior target probability map of the curve-shape area, generated by Parzen windows with Gaussian kernels, is approximated by the 1-D Gaussian mixture model (GMM) in order to extract some high-value curve segments corresponding to Gaussian components. Based on the parameterized curve segments from GMM, the self-organizing map (SOM) neural network is then established to achieve the coverage search. The step of winner neuron selection in SOM will prioritize and allocate the curve segments to UAVs, with the comprehensive consideration of multiple evaluation factors and allocation balance. The following step of neuron weight update will plan the UAV paths under the constraint of maneuverability and obstacle avoidance, using the modified Dubins guidance vector field. Finally, the good performance of GMM-SOM is evaluated on a coastline map.

We study the problem of asynchronous rendezvous search with multiple mobile agents (robots) in the plane. The goal of the robots is to meet at a location in the environment which is not determined in advance as quickly as possible. They do not know the initial locations of each other or their own initial locations. Moreover, the initial distance between any pair of robots is also unknown. Therefore, the problem must be solved using an online approach. We study a new variant of the rendezvous search problem which we call m -RP. The objective of m -RP is exactly m n of n robots to meet. n is the total number of robots in the environment, m is the number of robots that are active and want to meet, and k = n − m is the number robots that are waiting and not required to meet. None of the n robots knows m or k . We consider the symmetric version of the problem. In this version, the robots must execute the same strategy. For this problem, we present the algorithm m -RENDEZVOUS and show that its competitive ratio is O ( ( d k ) / R ) , where d is the initial distance between the furthest robot pair in the environment and R is the communication range of each robot. m -RENDEZVOUS utilizes the waiting robots in the environment to break the symmetry, and hence to achieve deterministic rendezvous. In addition to the theoretical results, we also verify the performance of our algorithm through extensive simulations.

IEEE Keyword search problem has been widely studied to retrieve relevant substructures from graphs. However, existing approaches aim at finding compact trees/subgraphs containing the keywords, and ignore density to evaluate how strongly and stablely the keyword nodes are connected. In this paper, we study the problem of finding cohesive subgraph containing query keywords with high density and compactness, and formulate it as minimal dense truss search problem based on k-truss model. However, unlike ${k}$-truss based community search that can be efficiently done by local search from a given set of nodes, this problem is nontrivial as the keyword nodes to be included in the retrieved substructure is previously unknown. To tackle this problem, we first design a novel hybrid KT-Index that keeps the keyword and truss information compactly to support efficient search of dense truss with the maximum trussness ${G_{den}}$. Then, we develop a novel refinement approach to extract minimal dense truss from ${G_{den}}$, by checking each node at most once based on the anti-monotonicity property of k-truss, together with several optimization strategies including batch based deletion, early-stop based deletion, and local exploration. Extensive experimental studies on real-world networks validated the effectiveness and efficiency of our approaches.

In recent years, with the rapid development of social networks, the scale of social networks has grown dramatically. The relationship of people in social networks plays a key role in information propagation. Due to its huge social value, the issue of influence maximization has attracted more and more researchers’ attention. With the increase of network scale, the running time has become a key issue. To resolve it, many heuristic algorithms for the influence maximization problem has been proposed. These algorithms have been verified to be be very efficient, but their effectiveness cannot be guaranteed for many networks in the real word. Therefore, it has become highly urgent to choose an influence maximization algorithm that is simultaneously efficient and effective. This paper proposes an algorithm based on the greedy algorithm framework, which converts the influence maximization problem into a maximum inner product search problem. This method can complete the selection of seed nodes for large-scale networks in nearly linear time. Compared with other algorithms, this algorithm has more application value.

Semantic GP is a promising branch of GP that introduces semantic awareness during genetic evolution to improve various aspects of GP. This paper presents a new Semantic GP approach based on Dynamic Target (SGP-DT) that divides the search problem into multiple GP runs. The evolution in each run is guided by a new (dynamic) target based on the residual errors of previous runs. To obtain the final solution, SGP-DT combines the solutions of each run using linear scaling. SGP-DT presents a new methodology to produce the offspring that does not rely on the classic crossover. The synergy between such a methodology and linear scaling yields final solutions with low approximation error and computational cost. We evaluate SGP-DT on eleven well-known data sets and compare with $$\epsilon$$ -lexicase, a state-of-the-art evolutionary technique, and seven Machine Learning techniques. SGP-DT achieves small RMSE values, on average 23.19% smaller than the one of $$\epsilon$$ -lexicase. Tuning SGP-DT ’s configuration greatly reduces the computational cost while still obtaining competitive results.

The graph search problem is the problem of searching for a mobile evader in a dark cave by a group of mobile searchers, where the cave is modeled by an undirected connected graph G. An offline and centralized version of the graph search problem is formalized as a pebble game on G, and has been extensively investigated under the name of the edge search problem. By es ( G ) we denote the number of pebbles (i.e., searchers) necessary and sufficient to edge search G. An online and distributed version of the graph search problem is the graph search by mobile agents. We investigate the graph search problem under a weaker online and distributed setting called the dark cave model, which models the dark cave more directly than a graph; G is completely anonymous, and we do not introduce any artificial facilities in G such as a port numbering, a homebase, and a whiteboard. Instead, we assume that the searchers can exchange information wherever they meet. Under the dark cave model, we propose a search algorithm OLSEARCH for es ( G ) searchers, and show that it is optimal in terms of the number of searchers.

We consider the surveillance target search problem for the case in which the arrival of targets is modeled by a spatio-temporal Poisson point process and the information measurement system scope is inhomogeneous. A probabilistic description of the inhomogeneity subdomains is given, and patterns of taking them into account when forming the a posteriori intensity density of a random point process are revealed. A search control law synthesis procedure for the case of an inhomogeneous scope is defined. An example is given.