A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is
To address the slow model re-training when data distribution shifts, Doraemon caches the previously-trained models and incrementally fine-tunes them for similar access patterns and data distribution. ... However, the current learned indexes assume the data distribution is relatively static and the access pattern is uniform, while real-world scenarios consist of skew query distribution and evolving data ... For the issue of model retraining, our insight is that the same model architecture can be reused for similar data distribution and access pattern. ...arXiv:1902.00655v1 fatcat:oeku72ktmjhlpeb5ts3a43gyk4
To achieve this, we first give the best strategy for an adversary to overload the system, and then prove that the cache size is lower bounded by O(n log log n/ log d), where n is the number of back-end ... Since log log n/ log d < 2 holds for almost all the current clusters (i.e., the number of back-end nodes n < 10 5 and the replication factor d ≥ 3), this result implies an O(n) lower bound on the required ... Formally, this adversarial access pattern can be described as a distribution S: S = (p 1 , p 2 , · · · , p m ) (1) where each p i denotes the fraction of queries for the ith key, p 1 + p 2 + ... + p m ...doi:10.1109/icdcsw.2013.24 dblp:conf/icdcsw/ChuGLCS13 fatcat:o7mb4oanhvba7kqlrmq4bdjp6u
IACR Cryptology ePrint Archive
In our main result, we present strong evidence that the leakage of the global key-equality pattern is inherent for any dynamic EMM construction with O(1) efficiency. ... Furthermore, we present stronger lower bounds that encrypted multi-maps leaking at most the decoupled key-equality pattern but are able to perform one of either the update or query operations in the plaintext ... Therefore, we must choose the hard distributions for each node v such that the decoupled equality pattern leakage is the same for the hard distribution of all nodes in the tree. ...dblp:journals/iacr/PatelPY19a fatcat:b6vnstxx5nfw5owu6zetx47ajq
(MapReduce-based) join execution for fast query responses. ... In this work we present H2RDF , a fully distributed RDF store that combines the MapReduce processing framework with a NoSQL distributed data store. ... The SP O index is responsible for triple patterns that have either bound subject or bound subject and predicate. ...doi:10.1145/2187980.2188058 dblp:conf/www/PapailiouKTK12 fatcat:6vdxhwgvhje23izc5aizmqpwxu
We adopt PROXIMUS, a recursive decomposition method for computing concise representations for binary data sets, to locally identify latent patterns in data distributed across peers. ... To enable efficient consolidation of patterns, we rely on distributed hash tables (DHT), commonly used in various applications in P2P networks. ... Vipin Kumar for his significant suggestions and comments on this work. This work is funded by the National Science Foundation grant CNS 0509387. ...doi:10.1016/j.jpdc.2007.06.003 fatcat:wy7v4i763vg3vmyabezkudhxl4
Pigeon's key pecking was reinforced with food in two experiments in which the correspondence between preference for starting one of two reinforced behavior patterns and the likelihood of finishing it subsequently ... For instance, preference for starting a pattern sometimes correlated negatively with the likelihood of subsequently completing it. ... two reinforced patterns, indexed by the lower bound of the shorter. ...doi:10.1901/jeab.1989.52-341 pmid:2584918 pmcid:PMC1339186 fatcat:qjs2epc7ibbu3c6mudxtl3ncke
This paper shows how a small, fast popularity-based front-end cache can ensure load balancing for an important class of such services; furthermore, we prove an O(n log n) lower-bound on the necessary cache ... We validate our analysis through simulation and empirical results running a key-value storage system on an 85-node cluster. ... the 85-node FAWN testbed used for this research. ...doi:10.1145/2038916.2038939 dblp:conf/cloud/FanLAK11 fatcat:rc3mcashk5g7zl4fga3cevm7hy
Lecture Notes in Computer Science
We present a block cipher structure and prove bounds on the resistance against differential and linear cryptanalysis. ... Assuming that the pairs are distributed over the trails according to a Poisson distribution, the expected number of pairs that, for a given key value, follow a differential trail with propagation probability ... for another value of the key. ...doi:10.1007/3-540-45325-3_20 fatcat:3cxoycjrozdy7av3gtczb47ufa
Lecture Notes in Computer Science
In order to set a minimal bound for the complexity of the Davies and Murphy attack, we want to get a maximal bound for the size of the deviation of the pattern of the output of the f function to the uniform ... We assume that there exists a pattern of n bits, in the output of f , so that the 2 n values this pattern may get are not equidistributed, for a given key K and uniformly distributed random input R. ...doi:10.1007/3-540-49649-1_13 fatcat:iswbvxpcpvfvnp6vb3epeaetsy
The graph data is only partially revealed to those users authorized for executing a query. ... Single triple patterns can be combined to query group patterns as they can be stated in SPARQL queries and allow to execute a wide range of SELECT and ASK queries. ... For example, a query of type -++ returns a set of subject URIs and requires a bound predicate and a bound object. A query is created from a basic key and two query patterns. ...dblp:conf/semweb/KastenSA013 fatcat:xvslkbuz6zc3rgg7644oot6bry
In addition, we give a comparison of the storage requirements of the key distribution patterns associated ... Summary: “A key distribution pattern is a combinatorial structure which provides a secure method of distributing secret keys among a number of participants in a cryptographic network. ...
For most major model types, we study implementations of key algorithms to determine common computational and runtime patterns. ... In this survey paper, we identify some of the key techniques employed in analytics both to serve as an introduction for the non-specialist and to explore the opportunity for greater optimizations for parallelization ... Distributed-memory task parallelism Memory-bound Data parallelism via SIMD or GPUs High-bandwidth, low-latency memory subsystem Pattern matching via FPGA Text Analytics Memory-bound Shared-and Distributed-memory ...doi:10.2200/s00678ed1v01y201511cac035 fatcat:jkjywe5rzzaupjwq5rjyavqxi4
Proceedings of the 2010 Workshop on Parallel Programming Patterns - ParaPLoP '10
We would like to especially thank Ralph Johnson, and Kurt Kuetzer for their input and encouragement. ... In fact, this minimal bound is asymptotically correct for the majority of distributions, unless the keys are mostly sorted initially. ... Further, the minimal communication bandwidth bound for a randomized distribution is T heta(n). ...doi:10.1145/1953611.1953621 fatcat:msddlq5jdbaehc22pimixtggdm
An experimental evaluation on distributed nodes shows that our model is well suited to provide time bounds for requests following typical workload patterns and that a prioritized extension can increase ... We formally define the pattern of the workload on our real-time DHT and use queuing theory to stochastically derive the time bound for response times of these lookup requests. ... RELATED WORK Distributed hash tables are well known for their performance and scalability for looking up keys on a node that stores associated values in a distributed environment. ...doi:10.1109/rtcsa.2014.6910537 dblp:conf/rtcsa/QianMX14 fatcat:42n672tfwvhepmtxbcivcvh4ga
Descriptive patterns are a key concept for inductive inference of pattern lan- gauges. ... This paper gives bounds on the tolerable amount of change (shift) of the underlying distribution such that learning is still possible in a PAC-like sense. ...
« Previous Showing results 1 — 15 out of 739,264 results