Diversity Forests: Using Split Sampling to Enable Innovative Complex Split Procedures in Random Forests

The diversity forest algorithm is an alternative candidate node split sampling scheme that makes innovative complex split procedures in random forests possible. While conventional univariable, binary splitting suffices for obtaining strong predictive performance, new complex split procedures can help tackling practically important issues. For example, interactions between features can be exploited effectively by bivariable splitting. With diversity forests, each split is selected from a candidate split set that is sampled in the following way: for $$l = 1, \dots , {nsplits}$$ l = 1 , ⋯ , nsplits : (1) sample one split problem; (2) sample a single or few splits from the split problem sampled in (1) and add this or these splits to the candidate split set. The split problems are specifically structured collections of splits that depend on the respective split procedure considered. This sampling scheme makes innovative complex split procedures computationally tangible while avoiding overfitting. Important general properties of the diversity forest algorithm are evaluated empirically using univariable, binary splitting. Based on 220 data sets with binary outcomes, diversity forests are compared with conventional random forests and random forests using extremely randomized trees. It is seen that the split sampling scheme of diversity forests does not impair the predictive performance of random forests and that the performance is quite robust with regard to the specified nsplits value. The recently developed interaction forests are the first diversity forest method that uses a complex split procedure. Interaction forests allow modeling and detecting interactions between features effectively. Further potential complex split procedures are discussed as an outlook.

Head Node Selection Algorithm in Cloud Computing Data Center

Cloud computing provides multiple services such as computational services, data processing, and resource sharing through multiple nodes. These nodes collaborate for all prementioned services in the data center through the head/leader node. This head node is responsible for reliability, higher performance, latency, and deadlock handling and enables the user to access cost-effective computational services. However, the optimal head nodes’ selection is a challenging problem due to consideration of resources such as memory, CPU-MIPS, and bandwidth. The existing methods are monolithic, as they select the head nodes without taking the resources of the nodes. Still, there is a need for the candidate node which can be selected as a head node in case of head node failure. Therefore, in this paper, we proposed a technique, i.e., Head Node Selection Algorithm (HNSA), for optimal head node selection from the data center, which is based on the genetic algorithm (GA). In our proposed method, there are three modules, i.e., initial population generation, head node selection, and candidate node selection. In the first module, we generate the initial population by randomly mapping the task on different servers using a scheduling algorithm. After that, we compute the overall cost and the cost of each node based on resources. In the second module, the best optimal nodes are selected as a head node by applying the genetic operations such as crossover, mutation, and fitness function by considering the available resources. In the selected optimal nodes, one node is chosen as a head node and the other is considered as a candidate node. In the third module, the candidate node becomes the head node in the case of head node failure. The proposed method HNSA is compared against the state-of-the-art algorithms such as Bees Life Algorithm (BLA) and Heterogeneous Earliest Finished Time (HEFT). The simulation analysis shows that the proposed HNSA technique performs better in terms of execution time, memory utilization, service level sgreement (SLA) violation, and energy consumption.

Design and Comparison of LEACH and Improved Centralized LEACH in Wireless Sensor Network

A WSN consists of a setup of sensor nodes/motes which perceives the environment under monitoring, and transfer this information through wireless links to the Base Station (BS) or sink. The sensor nodes can be heterogeneous or homogeneous and can be mobile or stationary. The data gathered is forwarded through single/multiple hops to the BS/sink. In this paper, propose improvements to LEACH routing protocol to reduce energy consumption and extend network life. LEACH Distance Energy (LEACH-DE) not only selects the cluster head node by considering that the remaining energy of the node is greater than the average remaining energy level of the nodes in the network, but also selects the cluster head node parameters based on the geometric distance between the candidate node and the BS. The simulation results show that the algorithm proposed in this work is superior to LEACH and LEACH-C (Centralized) in terms of energy saving and extending the lifetime of wireless sensor networks.

Towards an efficient weighted random walk domination

In this paper, we propose and study a new problem called the weighted random walk domination. Given a weighted graph G ( V, E ) and a budget B of the weighted random walk, it aims to find a k -size set S , which can minimize the total costs of the remaining nodes to access S through the weighted random walk, which is bounded by B. This problem is critical to a range of real-world applications, such as advertising in social networks and telecommunication base station selection in wireless sensor networks. We first present a dynamic programming based greedy method (DpSel) as a baseline. DpSel is time-consuming when | V | is huge. Thus, to overcome this drawback, we propose a matrix-based greedy method (MatrixSel), which can reduce the computation cost greatly. To further accelerate MatrixSel, we propose a BoundSel approach to reduce the number of the gain computations in each candidate selection by proactively estimating the upper bound of the marginal gain of the candidate node. Notably, all methods can achieve an approximation ratio of (1 - 1/ e ). Experiments on real datasets have been conducted to verify the efficiency, effectiveness, memory consumption and scalability of our methods.

Energy-Efficient Spatial Query-Centric Geographic Routing Protocol in Wireless Sensor Networks

In data-centric wireless sensor networks (WSNs), sensing data have a high time–space correlation. Most queries are spatial and used to obtain data in a defined region. Geographic routing (GR) protocols are the optimal choice for routing spatial queries. However, several drawbacks still exist in GRs, and these the include premature death of nodes and communication latency, which result in reduced network life and query efficiency. A new clustering GR protocol called quadtree grid (QTGrid) was proposed in this study to save energy and improve spatial query efficiency. First, the monitoring area was logically divided into clusters by a quadtree structure, and each grid’s location was encoded to reduce the memory overhead. Second, cluster head (CH) nodes were selected based on several metrics, such as distance from the candidate node to the grid center and adjacent CHs and residual energy. Third, the next-hop routing node was selected depending on the residual energy of the candidate node and its distance to the sink node. Lastly, a lossless data aggregation algorithm and a flexible spatial query algorithm were adopted to reduce the transmission of redundant data and meet the application requirements, respectively. Simulation results showed that compared with three related protocols, QTGrid has lower energy consumption and higher spatial query efficiency and is more suitable for large-scale WSN spatial query application scenarios.

ZBTB16 and Metabolic Syndrome: a Network Perspective

Metabolic syndrome is a prevalent, complex condition. The search for genetic determinants of the syndrome is currently undergoing a paradigm enhancement by adding systems genetics approaches to association studies. We summarize the current evidence on relations between an emergent new candidate, zinc finger and BTB domain containing 16 (ZBTB16) transcription factor and the major components constituting the metabolic syndrome. Information stemming from studies on experimental models with altered Zbtb16 expression clearly shows its effect on adipogenesis, cardiac hypertrophy and fibrosis, lipid levels and insulin sensitivity. Based on current evidence, we provide a network view of relations between ZBTB16 and hallmarks of metabolic syndrome in order to elucidate the potential functional links involving the ZBTB16 node. Many of the identified genes interconnecting ZBTB16 with all or most metabolic syndrome components are linked to immune function, inflammation or oxidative stress. In summary, ZBTB16 represents a promising pleiotropic candidate node for metabolic syndrome.

Secure and Distributed On-Demand Randomized Routing in WSN

Security and energy efficiency is of paramount importance in a wireless sensor network. This is due to their vulnerabledeployment conditions and battery based power. This paper presents a secure and distributed algorithm that generatesroutes on-demand in a wireless sensor network. Dynamic route generation is facilitated by PSO, a metaheuristictechnique. Current network traffic in that route and charge contained in the candidate node are used as evaluationparameters along with the node distance, hence a huge reduction in the packet loss was observed. Experiments wereconducted and it was observed that the proposed algorithm exhibits very low selection overhead and also providesdistributed routs, which eventually lead to prolonged network lifetime.