全文获取类型
收费全文 | 38篇 |
免费 | 3篇 |
专业分类
化学工业 | 3篇 |
水利工程 | 1篇 |
无线电 | 3篇 |
一般工业技术 | 2篇 |
自动化技术 | 32篇 |
出版年
2023年 | 1篇 |
2020年 | 1篇 |
2019年 | 1篇 |
2018年 | 2篇 |
2017年 | 1篇 |
2016年 | 2篇 |
2015年 | 1篇 |
2014年 | 2篇 |
2013年 | 1篇 |
2012年 | 1篇 |
2011年 | 2篇 |
2010年 | 2篇 |
2009年 | 1篇 |
2008年 | 1篇 |
2007年 | 2篇 |
2006年 | 4篇 |
2005年 | 1篇 |
2004年 | 3篇 |
2003年 | 1篇 |
2001年 | 1篇 |
1998年 | 1篇 |
1997年 | 2篇 |
1996年 | 1篇 |
1995年 | 1篇 |
1994年 | 3篇 |
1988年 | 1篇 |
1979年 | 1篇 |
排序方式: 共有41条查询结果,搜索用时 15 毫秒
21.
Zeyi?WenEmail author Rui?Zhang Kotagiri?Ramamohanarao Li?YangEmail author 《World Wide Web》2018,21(2):261-287
Support Vector Machine (SVM) regression is an important technique in data mining. The SVM training is expensive and its cost is dominated by: (i) the kernel value computation, and (ii) a search operation which finds extreme training data points for adjusting the regression function in every training iteration. Existing training algorithms for SVM regression are not scalable to large datasets because: (i) each training iteration repeatedly performs expensive kernel value computations, which is inefficient and requires holding the whole training dataset in memory; (ii) the search operation used in each training iteration considers the whole search space which is very expensive. In this article, we significantly improve the scalability and efficiency of SVM regression by exploiting the high performance of Graphics Processing Units (GPUs) and solid state drives (SSDs). Our key ideas are as follows. (i) To reduce the cost of repeated kernel value computations and avoid holding the whole training dataset in the GPU memory, we precompute all the kernel values and store them in the CPU memory extended by the SSD; together with an efficient strategy to read the precomputed kernel values, reusing precomputed kernel values with an efficient retrieval is much faster than computing them on-the-fly. This also alleviates the restriction that the training dataset has to fit into the GPU memory, and hence makes our algorithm scalable to large datasets, especially for large datasets with very high dimensionality. (ii) To enhance the performance of the frequently used search operation, we design an algorithm that minimizes the search space and the number of accesses to the GPU global memory; this optimized search algorithm also avoids branch divergence (one of the causes for poor performance) among GPU threads to achieve high utilization of the GPU resources. Our proposed techniques together form a scalable solution to the SVM regression which we call SIGMA. Our extensive experimental results show that SIGMA is highly efficient and can handle very large datasets which the state-of-the-art GPU-based algorithm cannot handle. On the datasets of size that the state-of-the-art GPU-based algorithm can handle, SIGMA consistently outperforms the state-of-the-art GPU-based algorithm by an order of magnitude and achieves up to 86 times speedup. 相似文献
22.
Oleksii Vedernikov Lars Kulik Kotagiri Ramamohanarao 《Journal of Location Based Services》2020,14(1):1-27
ABSTRACT Hitchhiking is a travel mode characterised by unpredictable travel times involving several possible combinations of lifts on roads. In this paper, we formulate a hitchhiker’s problem and develop a time-dependent stochastic route planning algorithm for hitchhikers. Namely, we introduce a concept of the stochastic time-dependent hitchhiking graph to find hitchhiking strategies with the least expected travel time or maximised reliability. We introduce various heuristics to prune the original hitchhiking graph to improve computational efficiency. We provide a complexity analysis of the problem and evaluate the proposed solution on real-world networks of several countries. 相似文献
23.
Variam Ullas Jeankumar Janupally Renuka Sonali Kotagiri Shalini Saxena Shruti Singh Kakan Jonnalagadda Padma Sridevi Swapna Yellanki Pushkar Kulkarni Prof. Perumal Yogeeswari Prof. Dharmarajan Sriram 《ChemMedChem》2014,9(8):1850-1859
In this study, we explored the pharmaceutically underexploited mycobacterial gyrase ATPase (GyrB) domain as a template for a structure‐based virtual screening of our in‐house (BITS Pilani) compound collection to discover new inhibitors targeting Mycobacterium tuberculosis (M.tb.) The hit identified was further customized by using a combination of molecular docking and medicinal chemistry strategies to obtain an optimized analogue displaying considerable in vitro enzyme efficacy and bactericidal properties against the M.tb. H37Rv strain. The binding affinity of the ligand toward the GyrB domain was reascertained by differential scanning fluorimetry experiments. Further evaluation of the hERG toxicity (a major limitation among the previously reported N‐linked aminopiperidine analogues) indicated these molecules to be completely devoid of cardiotoxicity, a significant achievement within this class. 相似文献
24.
Evan P. Harris Kotagiri Ramamohanarao 《The VLDB Journal The International Journal on Very Large Data Bases》1996,5(1):64-84
A method of analysing join algorithms based upon the time required
to access, transfer and perform the relevant CPU-based operations
on a disk page is proposed. The costs of variations of several of
the standard join algorithms, including nested block, sort-merge,
GRACE hash and hybrid hash, are presented. For a given total
buffer size, the cost of these join algorithms depends on the
parts of the buffer allocated for each purpose. For example, when
joining two relations using the nested block join algorithm, the
amount of buffer space allocated for the outer and inner relations
can significantly affect the cost of the join. Analysis of
expected and experimental results of various join algorithms show
that a combination of the optimal nested block and optimal GRACE
hash join algorithms usually provide the greatest cost benefit,
unless the relation size is a small multiple of the memory size.
Algorithms to quickly determine a buffer allocation producing the
minimal cost for each of these algorithms are presented. When the
relation size is a small multiple of the amount of main memory
available (typically up to three to six times), the hybrid hash
join algorithm is preferable.
Edited by Masaru Kitsuregawa.
Received April 26, 1993 / Revised March 3, 1994 /
Accepted October 13, 1994 相似文献
25.
Distributed computing systems and applications are not only changing the face of computing, they're also continually changing the way we live, work, and interact as a society. In recent years, the demand for educational courses in advanced computing and networking has expanded rapidly. This has led to a huge demand for Internet-based distributed computing technologies (such as Web and grid services) and applications that virtualize geographically distributed resources to enable the creation of virtual enterprises, marketplaces, and service-oriented computing industries. Because current academic programs don't yet address the skill set required to meet these demands, we launched a Master of Engineering in distributed computing program at the University of Melbourne 相似文献
26.
C. Levy Author Vitae M. Perl Author Vitae S. Kotagiri Author Vitae 《Engineering Fracture Mechanics》2006,73(13):1814-1825
The influence of the Bauschinger effect (BE) on the three-dimensional, Mode I, stress intensity factor (SIF) distributions for arrays of longitudinal coplanar, surface cracks emanating from the bore of a fully or partially autofrettaged thick-walled cylinder is investigated. The SIFs for both “realistic” - Bauschinger effect dependent autofrettage (BEDA) and “ideal” - Bauschinger effect independent autofrettage (BEIA) are obtained and compared. The 3D analysis is performed via the finite element (FE) method and the submodeling technique, employing singular elements along the crack front. Both autofrettage residual stress fields, BEDA and BEIA, are simulated using an equivalent temperature field. The Bauschinger effect (BE) is found to significantly lower the beneficial stress intensity factor due to autofrettage, KIA, by up to 52%, as compared to the “ideal” autofrettage case. The reduction in KIA varies along the crack front with the maximum determined by the crack ellipticity, crack depth and crack separation distance. The detrimental influence of the BE increases as the crack density decreases and as crack depth decreases. For a partially autofrettaged cylinder, the influence of the BE is considerably reduced as the level of overstrain becomes smaller. Furthermore, the results indicate that in certain situations crack density and crack ellipticity have opposing effects on the autofrettage SIF. 相似文献
27.
We present an efficient evaluation technique for modularly stratified deductive database programs for which the local strata level mappings are known at compile time. We present an important subclass of these programs (called EMS-programs) in which one can easily express problems, such as shortest distance, company ownership, bill of materials, and preferential vote counting. Programs written in this style have an easy-to-understand semantics and can be efficiently computed. Another important virtue of these programs is that their modular-stratification properties are independent of the extensional database 相似文献
28.
M. N. Halgamuge H. L. Vu K. Ramamohanarao M. Zukerman 《International Journal of Communication Systems》2011,24(3):363-383
This paper proposes a new approach for performance evaluation and comparison between existing handoff algorithms taking into consideration signal levels, call dropping, and handoff cost. Using the new approach, existing handoff algorithms are then compared in terms of signal quality and number of handoffs required to achieve a desired overall signal quality. We also provide in this paper a method to estimate handoff cost and to optimize handoff sequences for retrial (where repeated call attempts are made after a call is lost) and non‐retrial models based on the proposed approach. We observe that the Threshold with Hysteresis method performs better than other known methods including the one used in the GSM standard. Our results indicate that the Threshold with 4 dB Hysteresis method performs well for urban areas although with a high dropping probability, whereas the Threshold with 6 dB Hysteresis method suits for suburban areas with a low dropping probability. We find that handoff sequences obtained by existing handoff methods are less efficient than the optimal handoff sequence given in the paper by a margin of 29–45% for retrial model, and by 34–77% for non‐retrial model. The paper also suggests some specific parameter values to improve the performance of currently used handoff methods based on our findings. Copyright © 2010 John Wiley & Sons, Ltd. 相似文献
29.
Xindong?WuEmail author Philip?S.?Yu Gregory?Piatetsky-Shapiro Nick?Cercone T.?Y.?Lin Ramamohanarao?Kotagiri Benjamin?W.?Wah 《Knowledge and Information Systems》2003,5(2):248-261
At the 2001 IEEE International Conference on Data Mining in San Jose, California,
on November 29 to December 2, 2001, there was a panel discussion on how data
mining research meets practical development. One of the motivations for organizing the
panel discussion was to provide useful advice for industrial people to explore their directions
in data mining development. Based on the panel discussion, this paper presents
the views and arguments from the panel members, the Conference Chair and the Program
Committee Co-Chairs. These people as a group have both academic and industrial
experiences in different data mining related areas such as databases, machine learning,
and neural networks. We will answer questions such as (1) how far data mining is from
practical development, (2) how data mining research differs from practical development,
and (3) what are the most promising areas in data mining for practical development. 相似文献
30.
Zoltan Somogyi Kotagiri Ramamohanarao Jayen Vaghani 《International journal of parallel programming》1988,17(3):207-257
We present the first backtracking algorithm for stream AND-parallel logic programs. It relies on compile-time knowledge of
the dataflow graph of each clause to let it figure out efficiently which goals to kill or restart when a goal fails. This
crucial information, which we derive from mode declarations, was not available at compile-time in any previous stream AND-parallel
system.
We show that modes can increase the precision of the backtracking algorithm, though our algorithm allows this precision to
be traded off against overhead on a procedure-by-procedure and call-by-call basis. The modes also allow our algorithm to handle
efficiently programs that manipulate partially instantiated data structures and an important class of programs with circular
dependency graphs. On code that does not need backtracking, the efficiency of our algorithm approaches that of the committed-choice
languages; on code that does need backtracking its overhead is comparable to that of the independent AND-parallel backtracking
algorithms. 相似文献