Filters








126,075 Hits in 8.3 sec

Improving branch-and-cut performance by random sampling

Matteo Fischetti, Andrea Lodi, Michele Monaci, Domenico Salvagnin, Andrea Tramontani
2015 Mathematical Programming Computation  
We discuss the variability in the performance of multiple runs of branch-and-cut Mixed Integer Linear Programming solvers, and we concentrate on the one deriving from the use of different optimal bases  ...  We propose a new algorithm exploiting more than one of those bases and we show that different versions of the algorithm can be used to stabilize and improve the performance of the solver.  ...  Acknowledgements This research was partially supported by MiUR, Italy (PRIN project "Mixed-Integer Nonlinear Optimization: Approaches and Applications") and by the University of Padova (Progetto di Ateneo  ... 
doi:10.1007/s12532-015-0096-0 fatcat:jnrgly42rrhl5n3fvizgnpekta

Extended Isolation Forest [article]

Sahand Hariri, Matias Carrasco Kind, Robert J. Brunner
2018 arXiv   pre-print
We show that the consistency and reliability of the algorithm is much improved using this method by looking at the variance of scores of data points distributed along constant score lines.  ...  Second, which is the preferred way, is to allow the slicing of the data to use hyperplanes with random slopes. This approach results in improved score maps.  ...  For the extended case, the selection of the branch cuts still only requires two pieces of information, but they are: 1) a random slope for the branch cut, and 2) a random intercept for the branch cut which  ... 
arXiv:1811.02141v1 fatcat:fecjcgc72rbmlj53mq5iieavze

Exploring Data Augmentation for Multi-Modality 3D Object Detection [article]

Wenwei Zhang, Zhe Wang, Chen Change Loy
2021 arXiv   pre-print
In addition, considering occlusions, a point in different modalities may be occupied by different objects, making augmentations such as cut and paste non-trivial for multi-modality detection.  ...  We further present Multi-mOdality Cut and pAste (MoCa), which simultaneously considers occlusion and physical plausibility to maintain the multi-modality consistency.  ...  The lower value of PKL means that the predictions are more similar to ground truth and are more suitable for planning.  ... 
arXiv:2012.12741v2 fatcat:yslkh6yt2rd5nmmnmeawhnetaq

Extended Isolation Forest with Randomly Oriented Hyperplanes

Sahand Hariri, Matias Carrasco Kind, Robert J. Brunner
2019 IEEE Transactions on Knowledge and Data Engineering  
These maps suffer from artifacts generated by the criteria for branching operation of the binary tree. We explain this problem in detail and demonstrate the mechanism by which it occurs visually.  ...  We show that the robustness of the algorithm is much improved using this method by looking at the variance of scores of data points distributed along constant level sets.  ...  MCK's work has been supported by Grant projects NSF AST 07-15036 and NSF AST 08-13543.  ... 
doi:10.1109/tkde.2019.2947676 fatcat:37fjgxhavvdklaeaiolnctrnna

Object classification at the Nearby Supernova Factory

S. Bailey, C. Aragon, R. Romano, R. C. Thomas, B. A. Weaver, D. Wong
2008 Astronomical Notes - Astronomische Nachrichten  
In comparison to simple threshold cuts, more sophisticated methods such as boosted decision trees, random forests, and support vector machines provide dramatically better object discrimination: we reduced  ...  the number of nonsupernova candidates by a factor of 10 while increasing our supernova identification efficiency.  ...  DE-AC02-05CH11231, and by a grant from the Gordon & Betty Moore Foundation.  ... 
doi:10.1002/asna.200710932 fatcat:zsj2dhlf7rgcdn7y4gxwk4ll5q

An Efficient RRT Algorithm for Motion Planning of Live-Line Maintenance Robots

Jiabo Feng, Weijun Zhang
2021 Applied Sciences  
In the process of expanding the tree, by removing small branches, the tree can fully wiring in the planning space with a small number of nodes.  ...  Through dynamic sampling and pruning, the growth of trees in undesired areas is reduced and undesired planning results are avoided.  ...  Some branches of T cut may go through the dynamic obstacle. Detecting these branches consumes many resources, so no detection is performed.  ... 
doi:10.3390/app112210773 fatcat:mgfwviannrfbdhawpn4jg7m6km

Batch DropBlock Network for Person Re-identification and Beyond [article]

Zuozhuo Dai and Mingqiang Chen and Xiaodong Gu and Siyu Zhu and Ping Tan
2019 arXiv   pre-print
In this paper, we propose the Batch DropBlock (BDB) Network which is a two branch network composed of a conventional ResNet-50 as the global branch and a feature dropping branch.  ...  The network then concatenates features from both branches and provides a more comprehensive and spatially distributed feature representation.  ...  Adding the global branch could further improve the performance.  ... 
arXiv:1811.07130v2 fatcat:syqyrouhsnczfp3wi57jupwwsm

Java Unit Testing Tool Competition

Xavier Devroey, Sebastiano Panichella, Alessio Gambi
2020 Proceedings of the IEEE/ACM 42nd International Conference on Software Engineering Workshops  
This paper describes our methodology and statistical analysis of the results, presents the results achieved by the contestant tools and highlights the challenges we faced during the competition.  ...  CCS CONCEPTS • Software and its engineering → Search-based software engineering; Automatic programming; Software testing and debugging.  ...  This research was partially funded by the EU Horizon 2020 ICT-10-2016-RIA âĂIJSTAMPâĂİ project (No.731529).  ... 
doi:10.1145/3387940.3392265 dblp:conf/icse/DevroeyPG20 fatcat:5yg2wyk2z5cdtmlm2w63u3vtnu

Learning to Select Cuts for Efficient Mixed-Integer Programming [article]

Zeren Huang, Kerong Wang, Furui Liu, Hui-ling Zhen, Weinan Zhang, Mingxuan Yuan, Jianye Hao, Yong Yu, Jun Wang
2021 arXiv   pre-print
To measure the quality of the candidate cuts, a scoring function, which takes the instance-specific cut features as inputs, is trained and applied in cut ranking and selection.  ...  In this paper, we propose a data-driven and generalizable cut selection approach, named Cut Ranking, in the settings of multiple instance learning.  ...  Acknowledgements Weinan Zhang is supported by "New Generation of AI 2030" Major Project (2018AAA0100900) and National Natural Science Foundation of China (62076161).  ... 
arXiv:2105.13645v4 fatcat:nh5h6sds3zgt3eo5qbwqkyfq5a

Combining simulated annealing with local search heuristics

Olivier C. Martin, Steve W. Otto
1996 Annals of Operations Research  
For the traveling salesman problem with randomly distributed cities in a square, the procedure improves on 3-opt by 1.6%, and on Lin-Kernighan local search by 1.3%.  ...  We have tested this meta-heuristic for the traveling salesman and graph partitioning problems. Tests on instances from public libraries and random ensembles quantify the power of the method.  ...  Johnson, and H. Simon for help, criticism, and discussions. The comments of the referees were very helpful and allowed us to greatly improve the manuscript.  ... 
doi:10.1007/bf02601639 fatcat:mtlpomzmfrehppka5ghzwho5hm

Extending rapidly-exploring random trees for asymptotically optimal anytime motion planning

Yasin Abbasi-Yadkori, Joseph Modayil, Csaba Szepesvari
2010 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems  
The state-of-the-art solution to this problem is the rapidlyexploring random tree (RRT) algorithm that quickly finds a feasible solution.  ...  We review the existing approaches to this problem and find no algorithms that both quickly find feasible solutions and also eventually approach optimal solutions with additional time.  ...  If RRT tries to extend b1 by sampling more states, b2 will absorb samples near the goal and prevent b1 from growing towards the goal. In Figure 8(a) , a sub-branch of b2 has blocked b1.  ... 
doi:10.1109/iros.2010.5650614 dblp:conf/iros/Abbasi-YadkoriMS10 fatcat:7cds6othqnbxng5vnnieqnusdi

Geometry Feature Extraction of Shorea Leaf Venation Based on Digital Image and Classification Using Random Forest

Ishak Ariawan
2022 International Journal of Computing and Digital Systems  
Also, optimum accuracy was attained at 91.90%, with cut-off training and testing data of 90:10, by analyzing 1000 single trees.  ...  As a consequence, precise and rapid identification system is required to differentiate the sample at the seedling phase.  ...  Acknowledgment We greatly appreciate the research support from both Department of Computer Science, Faculty of Mathematics and Natural Sciences and Department of Silviculture, Faculty of Forestry and Environment  ... 
doi:10.12785/ijcds/110111 fatcat:6iseperg6jdilgq5nctsc2q434

Java Unit Testing Tool Competition - Seventh Round

Fitsum Kifetew, Xavier Devroey, Urko Rueda
2019 2019 IEEE/ACM 12th International Workshop on Search-Based Software Testing (SBST)  
Given the interesting findings of last year, we analyzed the effectiveness of the combined test suites generated by all competing tools and compared; results are confronted with the manual test suites  ...  of the projects, as well as those generated by the competing tools .  ...  Moreover, we sampled the mutants generated by PITest: we applied a random sampling of 33% for CUTs with more than 200 mutants, and a sampling of 50% for CUTs with more than 400 mutants.  ... 
doi:10.1109/sbst.2019.00014 dblp:conf/icse/KifetewDR19 fatcat:b5tsmc6u7bbatftqdvpz5dt4zq

Exploring the Impact of Negative Samples of Contrastive Learning: A Case Study of Sentence Embedding [article]

Rui Cao, Yihao Wang, Yuxin Liang, Ling Gao, Jie Zheng, Jie Ren, Zheng Wang
2022 arXiv   pre-print
This is typically achieved by maintaining a queue of negative samples during training.  ...  We add the prediction layer to the online branch to make the model asymmetric and together with EMA update mechanism of the target branch to prevent the model from collapsing.  ...  As shown in Table 5 , cut off and token shuffle do not improve, even slightly hurt the model's performance. Only the adversarial attack (FGSM) has slight improvement on the performance.  ... 
arXiv:2202.13093v3 fatcat:fl66si4otbbuhesmgio37r7ihe

Using diversification, communication and parallelism to solve mixed-integer linear programs

R. Carvajal, S. Ahmed, G. Nemhauser, K. Furman, V. Goel, Y. Shao
2014 Operations Research Letters  
We propose a framework using multiple branch-and-bound trees to solve MILPs while allowing them to share information in a parallel execution.  ...  Performance variability of modern mixed-integer programming solvers and possible ways of exploiting this phenomenon present an interesting opportunity in the development of algorithms to solve mixed-integer  ...  Acknowledgements This research has been supported in part by ExxonMobil Upstream Research Company, AFOSR and the School of Bussiness at Universidad Adolfo Ibáñez.  ... 
doi:10.1016/j.orl.2013.12.012 fatcat:34trkzdrx5ecxfwjg5hsda2ure
« Previous Showing results 1 — 15 out of 126,075 results