Filters








36,585 Hits in 7.9 sec

Naïve Bayesian filters for log file analysis: Despam your logs

R. W. Havens, B. Lunt, Chia-Chi Teng
2012 2012 IEEE Network Operations and Management Symposium  
Naïve Bayesian Spam Filters for Log File Analysis Russel W.  ...  For stages 2 and 3, log entries were tested with digits normalized to zeros, with words chained together to various lengths and one or all levels of word chains used together.  ...  After all, a word processor, a network driver, an operating system and a web server will have very different logging needs.  ... 
doi:10.1109/noms.2012.6211972 dblp:conf/noms/HavensLT12 fatcat:6ve6ei7ubvbghfwtma3crcbnbq

Learning robot grasping from 3-D images with Markov Random Fields

A. Boularias, O. Kroemer, J. Peters
2011 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems  
The empirical results show a significant improvement over methods that do not utilize the smoothness assumption and classify each point separately from the others.  ...  Our approach is based on Markov Random Fields (MRF), and motivated by the fact that points that are geometrically close to each other tend to have similar grasp success probabilities.  ...  The AMN model uses the log-linear function for representing a potential as a function of the features, i.e. log φ i (y) = w T n,y x i and log φ ij (y) = w T e,y x ij , where w n,y ∈ R dn are the weights  ... 
doi:10.1109/iros.2011.6048528 fatcat:3ouqyvgfqngldoyfg3q4og2wda

Inhaled nitric oxide for the adjunctive therapy of severe malaria: Protocol for a randomized controlled trial

Michael Hawkes, Robert O Opoka, Sophie Namasopo, Christopher Miller, Kevin E Thorpe, James V Lavery, Andrea L Conroy, W Conrad Liles, Chandy C John, Kevin C Kain
2011 Trials  
Endothelial activation plays a central role in the pathogenesis of severe malaria, of which angiopoietin-2 (Ang-2) has recently been shown to function as a key regulator.  ...  Nitric oxide (NO) is a major inhibitor of Ang-2 release from endothelium and has been shown to decrease endothelial inflammation and reduce the adhesion of parasitized erythrocytes.  ...  Time to recovery Among survivors (a subgroup of randomized participants), recovery times will be analysed by survival analysis (log-rank test).  ... 
doi:10.1186/1745-6215-12-176 pmid:21752262 pmcid:PMC3151218 fatcat:eu5aztfjyfekdf73hfpz3cqpuq

Smoothed Analysis of Dynamic Networks [article]

Michael Dinitz, Jeremy T. Fineman, Seth Gilbert, Calvin Newport
2015 arXiv   pre-print
We generalize the technique of smoothed analysis to distributed algorithms in dynamic network models.  ...  We prove that these bounds provide a spectrum of robustness when subjected to smoothing---some are extremely fragile (random walks), some are moderately fragile / robust (flooding), and some are extremely  ...  An O(n 2/3 log n/k 1/3 ) Upper Bound for General Networks We now show that flooding in every k-smoothed network will complete in O(n 2/3 log n/k 1/3 ) time, with high probability.  ... 
arXiv:1508.03579v1 fatcat:57fdzhu2srbxlbk6phab3ue7pq

Smoothed Analysis of Dynamic Networks [chapter]

Michael Dinitz, Jeremy Fineman, Seth Gilbert, Calvin Newport
2015 Lecture Notes in Computer Science  
Whereas in the traditional setting smoothing typically perturbs numerical input values, in our setting we define smoothing to perturb the network graph through the random addition and deletion of edges  ...  Whereas standard smoothed analysis studies the impact of small random perturbations of input values on algorithm performance metrics, dynamic graph smoothed analysis studies the impact of random perturbations  ...  An O(n 2/3 log n/k 1/3 ) Upper Bound for General Networks Next, we show that flooding in every k-smoothed network will complete in O(n 2/3 log n/k 1/3 ) time, with high probability.  ... 
doi:10.1007/978-3-662-48653-5_34 fatcat:ap72iwnbtfcelpjxweevdgc2ky

Smoothed analysis of dynamic networks

Michael Dinitz, Jeremy T. Fineman, Seth Gilbert, Calvin Newport
2017 Distributed computing  
Whereas in the traditional setting smoothing typically perturbs numerical input values, in our setting we define smoothing to perturb the network graph through the random addition and deletion of edges  ...  Whereas standard smoothed analysis studies the impact of small random perturbations of input values on algorithm performance metrics, dynamic graph smoothed analysis studies the impact of random perturbations  ...  An O(n 2/3 log n/k 1/3 ) Upper Bound for General Networks Next, we show that flooding in every k-smoothed network will complete in O(n 2/3 log n/k 1/3 ) time, with high probability.  ... 
doi:10.1007/s00446-017-0300-8 fatcat:5wlkxg7tubcv5nimwvqn6m3qnu

Bounding Communication Cost in Dynamic Load Balancing of Distributed Hash Tables [chapter]

Marcin Bienkowski, Miroslaw Korzeniowski
2006 Lecture Notes in Computer Science  
Our procedure requires O(log n) times more messages than any procedure maintaining the connectivity, even if the an oblivious adversary decides about the dynamics of the system.  ...  As a byproduct, we show how to compute a constant approximation of the current number of nodes n in the system, provided that we know an upper bound on log n.  ...  D is O(log n) in Chord [2] and O(log n/ log log n) in de Bruijn graph [5, 6] . In Subsection 2.1 we introduce a notion of weight of an interval.  ... 
doi:10.1007/11795490_29 fatcat:46zil2buqfeypobifm4u2bjpoq

A Dual Approach for Optimal Algorithms in Distributed Optimization over Networks [article]

César A. Uribe and Soomin Lee and Alexander Gasnikov and Angelia Nedić
2020 arXiv   pre-print
We study dual-based algorithms for distributed convex optimization problems over networks, where the objective is to minimize a sum ∑_i=1^mf_i(z) of functions over in a network.  ...  convex nor smooth.  ...  Doan, who made a lot very useful comments on the initial version of this text. This comments allows to repairs significant misprints. Funding The work of A. Nedić and C.A.  ... 
arXiv:1809.00710v3 fatcat:kpjafs6pjrc67afhqkqw6icsvy

Distributed MST: A Smoothed Analysis [article]

Soumyottam Chatterjee, Gopal Pandurangan, Nguyen Dinh Pham
2019 arXiv   pre-print
We present a distributed algorithm that, with high probability,[%s] computes an MST and runs in Õ(min{1/√(ϵ(n)) 2^O(√(log n)), D + √(n)}) rounds[%s] where ϵ is the smoothing parameter, D is the network  ...  To complement our upper bound, we also show a lower bound of Ω̃(min{1/√(ϵ(n)), D+√(n)}). We note that the upper and lower bounds essentially match except for a multiplicative 2^O(√(log n))(n) factor.  ...  There is a multicommodity routing algorithm on a random graph G(n, (log n)) that achieves congestion and dilation 2 O( √ log n) , and runs in time 2 O( √ log n) .  ... 
arXiv:1911.02628v1 fatcat:2m2nyawsjvfbrhshzwyqusvccu

Geographic Gossip: Efficient Aggregation for Sensor Networks [article]

Alexandros G. Dimakis, Anand D. Sarwate, Martin J. Wainwright
2006 arXiv   pre-print
In particular, for random geometric graphs, our algorithm computes the true average to accuracy 1/n^a using O(n^1.5√( n)) radio transmissions, which reduces the energy consumption by a √(n/ n) factor over  ...  For realistic sensor network model topologies like grids and random geometric graphs, the inefficiency of gossip schemes is caused by slow mixing times of random walks on those graphs.  ...  form E (n, 1/n a ) = O(n 3/2log n) and D(n, ǫ) = O(n 3/2 log 3/2 n).  ... 
arXiv:cs/0602071v1 fatcat:bx26uj4havcbnidvqell6nvvzm

Training (Overparametrized) Neural Networks in Near-Linear Time [article]

Jan van den Brand, Binghui Peng, Zhao Song, Omri Weinstein
2020 arXiv   pre-print
Very recently, this computational overhead was mitigated by the works of [ZMG19,CGH+19, yielding an O(mn^2)-time second-order algorithm for training two-layer overparametrized neural networks of polynomial  ...  Our result provides a proof-of-concept that advanced machinery from randomized linear algebra – which led to recent breakthroughs in 𝑐𝑜𝑛𝑣𝑒𝑥 𝑜𝑝𝑡𝑖𝑚𝑖𝑧𝑎𝑡𝑖𝑜𝑛 (ERM, LPs, Regression) – can be  ...  Suppose the width of the neural network satisfies m = Ω(max{λ −4 n 4 , λ −2 n 2 d log(16n/δ)}), then with probability 1 − δ over the random initialization of neural network and the randomness of the algorithm  ... 
arXiv:2006.11648v2 fatcat:b7dmwuurivetnafmbznqtghnum

On estimating the average degree

Anirban Dasgupta, Ravi Kumar, Tamas Sarlos
2014 Proceedings of the 23rd international conference on World wide web - WWW '14  
In this work we consider the problem of estimating the average degree of a large network using efficient random sampling, where the number of nodes is not known to the algorithm.  ...  While there has been a spate of recent work on estimating the number of nodes in a network, the edge-estimation question appears to be largely unaddressed.  ...  used by Smooth, Guess&Smooth executes at most log 2 (U/L) iterations, each with Θ(log(1/δ) + log log(U/L)) samples.  ... 
doi:10.1145/2566486.2568019 dblp:conf/www/DasguptaKS14 fatcat:drf6fumpvrejvjg4c2flru7bnu

Particle Smoothing Variational Objectives [article]

Antonio Khalil Moretti, Zizhao Wang, Luhuan Wu, Iddo Drori, Itsik Pe'er
2019 arXiv   pre-print
at the rate O(1/√(K)).  ...  Inspired by this work, we introduce Particle Smoothing Variational Objectives (SVO), a novel backward simulation technique and smoothed approximate posterior defined through a subsampling process.  ...  ∇ log Z + T t=2 T t ≥t+1 E ∇ w 1 t−1 Zt−1 · (w 1 t −Z t ) 2 2Z 2 t a 1 t−1 = 1 + O( 1 /K) 1 /K T t=1 E (∇ w 1 t Zt ) 2 + T t =t,t =1 T t=1 Var ∇ w 1 t Zt Var ∇ w 1 t Z t + O( T 2 /K 2 ) (16) where Z =  ... 
arXiv:1909.09734v1 fatcat:mjmmvwcp5jd47dvmxbmemgv6ia

Geographic gossip: efficient aggregation for sensor networks

A.G. Dimakis, A.D. Sarwate, M.J. Wainwright
2006 2006 5th International Conference on Information Processing in Sensor Networks  
In particular, for random geometric graphs, our algorithm computes the true average to accuracy 1/n a using O(n 1.5 √ log n) radio transmissions, which reduces the energy consumption by a q n log n factor  ...  For realistic sensor network model topologies like grids and random geometric graphs, the inefficiency of gossip schemes is caused by slow mixing times of random walks on those graphs.  ...  n, ) = O(n 3/2 log 3/2 n).  ... 
doi:10.1109/ipsn.2006.244081 fatcat:7z4oyj4s7zazxejchv6a3x747y

Geographic gossip

Alexandros G. Dimakis, Anand D. Sarwate, Martin J. Wainwright
2006 Proceedings of the fifth international conference on Information processing in sensor networks - IPSN '06  
In particular, for random geometric graphs, our algorithm computes the true average to accuracy 1/n a using O(n 1.5 √ log n) radio transmissions, which reduces the energy consumption by a q n log n factor  ...  For realistic sensor network model topologies like grids and random geometric graphs, the inefficiency of gossip schemes is caused by slow mixing times of random walks on those graphs.  ...  n, ) = O(n 3/2 log 3/2 n).  ... 
doi:10.1145/1127777.1127791 dblp:conf/ipsn/DimakisSW06 fatcat:pbkcu3uicnf3pancc4sz5q5rqu
« Previous Showing results 1 — 15 out of 36,585 results