Filters








22,799 Hits in 6.9 sec

Greedy Recommending Is Not Always Optimal [chapter]

Maarten van Someren, Vera Hollink, Stephan ten Hagen
2004 Lecture Notes in Computer Science  
Here we argue that in sequential recommending a series of normal, "greedy", recommendings is not always the strategy that minimises the number of steps in the search.  ...  Recommender systems help users to find objects or documents on web sites. In many cases it is not easy to know in advance by whom and for what purpose a web site will be used.  ...  Our main point is to demonstrate that greedy recommending is not always the optimal method for this problem.  ... 
doi:10.1007/978-3-540-30123-3_9 fatcat:x4xx6k3lrjhkri5wy62qbob4ve

Breaking out of the box of recommendations

Min Xie, Laks V.S. Lakshmanan, Peter T. Wood
2010 Proceedings of the fourth ACM conference on Recommender systems - RecSys '10  
In these contexts, there is a need for a system that can recommend top-k packages for the user to choose from.  ...  Because the problem of generating the top recommendation (package) is NP-complete, we devise several approximation algorithms for generating topk packages as recommendations.  ...  However, similar to Greedy-CR, it is obvious that Greedy-CR-Topk is not instance optimal.  ... 
doi:10.1145/1864708.1864739 dblp:conf/recsys/XieLW10 fatcat:iuj5andfh5enbjhmbhfbikrcey

Optimal Bayesian Recommendation Sets and Myopically Optimal Choice Query Sets

Paolo Viappiani, Craig Boutilier
2010 Neural Information Processing Systems  
Since recommendation set optimization is a simpler, submodular problem, this can greatly reduce the complexity of both exact and approximate (greedy) computation of optimal choice queries.  ...  We show that, under very general assumptions, the optimal choice query w.r.t. EVOI coincides with the optimal recommendation set, that is, a set maximizing the expected utility of the user selection.  ...  Not only is the optimal recommendation set problem somewhat easier computationally, it is submodular, admitting a greedy algorithm with approximation guarantees.  ... 
dblp:conf/nips/ViappianiB10 fatcat:nvdrr7lfnfh4vmuim5ykup2bji

A Randomized Algorithm for Maximizing the Diversity of Recommendations

K Alodhaibi, A Brodsky, G A Mihaila
2011 2011 44th Hawaii International Conference on System Sciences  
Second, we introduce a randomized algorithm, which is based on iterative relaxation of selections by the Greedy algorithm with an exponential probability distribution.  ...  the optimal solutions on cases run with the exhaustive algorithm in under 100 ms.  ...  As it stands, Greedy will always choose the case that has maximum distance to what has been chosen so far, but this may not provide the optimum solution toward the end.  ... 
doi:10.1109/hicss.2011.25 dblp:conf/hicss/AlodhaibiBM11 fatcat:y6cws2x4tvalvf33u5qrxgan24

Ordered Submodularity and its Applications to Diversifying Recommendations [article]

Jon Kleinberg, Emily Ryu, Éva Tardos
2022 arXiv   pre-print
A fundamental task underlying many important optimization problems, from influence maximization to sensor placement to content recommendation, is to select the optimal group of k items from a larger set  ...  Although extensions of submodularity to sequences have been proposed, none is designed to model settings where items contribute based on their position in a ranked list, and hence they are not able to  ...  So it is not always inherently better to rank i 1 before i 2 or i 2 before i 1 ; the optimal ordering is dependent on the context of the rest of the recommended list.  ... 
arXiv:2203.00233v1 fatcat:wehcqlotkjantllio3pbjim2sm

Optimizing Expert Systems: Heuristics for Efficiently Generating Low-Cost Information Acquisition Strategies

Michael V. Mannino, Vijay S. Mookerjee
1999 INFORMS journal on computing  
Our results indicate that the heuristics are very precise leading to near optimal results for greedy search and moderate search effort for optimal search.  ...  The heuristics are embedded into an informed search algorithm (based on AO*) that produces an optimal strategy and a greedy search algorithm.  ...  The second reason for investigating greedy search is that informed optimal search may not be possible for large rule bases.  ... 
doi:10.1287/ijoc.11.3.278 fatcat:h2kkcfhhwfaxhgzwugpcpjugym

Fighting Boredom in Recommender Systems with Linear Reinforcement Learning

Romain Warlop, Alessandro Lazaric, Jérémie Mary
2018 Neural Information Processing Systems  
A common assumption in recommender systems (RS) is the existence of a best fixed recommendation strategy.  ...  may outperform both fixed-action and contextual greedy policies.  ...  One strategy could be to apply a bandit algorithm to find the optimal genre and then always recommend movies of this genre.  ... 
dblp:conf/nips/WarlopLM18 fatcat:57q5c5qdarah7m5gz673oezv2m

Algorithmic Thinking - Greedy/Brute-Force/DynamicProgramming [chapter]

Raymond Xu
2020 Zenodo  
This article uses story telling approach and guide audience through a wedding shop recommendation scenario to explain the differences between Greedy algorithm, Brute Force and Dynamic Programming.  ...  What if we memorize the optimal choice of V given the budget of $140? Then we do not have to re-compute it later.  ...  Brute-Force Algorithm to Save the Falling Pants Will the Greedy Algorithm always work? What if the customer only has $320 budget instead of $360? Pick the most expensive tuxedo ($150)!  ... 
doi:10.5281/zenodo.4037386 fatcat:j36drc7ffzf45i5wwrqqszxqba

Contextual Exploration Using a Linear Approximation Method Based on Satisficing [article]

Akane Minami, Yu Kono, Tatsuji Takahashi
2021 arXiv   pre-print
To address this problem, we focus on the satisficing policy, which is a qualitatively different approach from that of existing optimization algorithms.  ...  However, the amount of exploration required for learning is often quite large.  ...  The optimal aspiration level of LinRS calculated from equation ( 4 ) is unlikely not always constant.  ... 
arXiv:2112.06452v1 fatcat:nt6mnjckqzdh5gcnjjdsdyhlma

Buyer to Seller Recommendation under Constraints [article]

Cheng Chen, Lan Zheng, Venkatesh Srinivasan, Alex Thomo, Kui Wu, Anthony Sukow
2014 arXiv   pre-print
On the other hand, we show that CAC-REC is NP-hard.  ...  We propose two approximate algorithms to solve CAC-REC and show that they achieve close to optimal solutions via comprehensive experiments using real-world datasets.  ...  Meanwhile, we also observe that the performance of GREEDY is always superior to that of SDP with rounding. Both SDP with rounding and GREEDY achieve close to optimal solutions.  ... 
arXiv:1406.0455v3 fatcat:gcdsbwlrhjb6la4dffbf7kyc4a

A Greedy-based Algorithm in Optimizing Student's Recommended Timetable Generator with Semester Planner

Khyrina Airin Fariza Abu Samah, Siti Qamalia Thusree, Ahmad Firdaus Ahmad Fadzil, Lala Septem Riza, Shafaf Ibrahim, Noraini Hasan
2022 International Journal of Advanced Computer Science and Applications  
Hence, this research aims to optimize the recommended semester planner, Timetable Generator using a greedy algorithm to increase student productivity.  ...  We calculate the priority task sequence to achieve the best optimal solution. The greedy algorithm can solve the optimization problem with the best optimal solution for each situation.  ...  CONCLUSION We presented a recommended semester planner using the optimization technique greedy optimization.  ... 
doi:10.14569/ijacsa.2022.0130146 fatcat:vwyso6qqefbhbpayivj5tia2fq

Stackelberg Strategic Guidance for Heterogeneous Robots Collaboration [article]

Yuhan Zhao, Baichuan Huang, Jingjin Yu, Quanyan Zhu
2022 arXiv   pre-print
With built-in tolerance of model uncertainty, the strategic guidance generated by our planning algorithm not only improves the overall efficiency in solving the rearrangement tasks, but is also robust  ...  The follower may not precisely execute the leader's recommended strategy if he is susceptible to uncertainties.  ...  The "no trust" in Fig. 5b means that the follower does not trust the leader's recommendation and selects the greedy strategy for manipulation.  ... 
arXiv:2202.01877v1 fatcat:ct65zt3egrcbrk3rh6zs2owvdu

Optimizing an Utility Function for Exploration / Exploitation Trade-off in Context-Aware Recommender System [article]

Djallel Bouneffouf
2014 arXiv   pre-print
This consists of optimizing a utility function represented by a linearized form of the probability distributions of the rewards of the clicked and the non-clicked documents already recommended.  ...  In this paper, we develop a dynamic exploration/ exploitation (exr/exp) strategy for contextual recommender systems (CRS).  ...  LINEARIZED -E-GREEDY To improve exploitation of the ε-greedy algorithm, we propose to use the linearization of clicked and not-clicks documents and the optimization of the utility function at the beginning  ... 
arXiv:1303.0485v2 fatcat:mtimh2qxmzgl7kcskwlotfdgxq

Show Me the Money: Dynamic Recommendations for Revenue Maximization [article]

Wei Lu, Shanshan Chen, Keqian Li, Laks V.S. Lakshmanan
2015 arXiv   pre-print
As recommendations can be seen as an information channel between a business and its customers, it is interesting and important to investigate how to make strategic dynamic recommendations leading to maximum  ...  As a result, most existing techniques are not explicitly built for revenue maximization, the primary business goal of enterprises.  ...  GlobalNo is always behind G-Greedy (about 10% to 30% loss), and so is SL-Greedy compared to RL-Greedy (about 1% to 6% behind).  ... 
arXiv:1409.0080v3 fatcat:hetbenmrtfgexm7ixs7wfpl3tu

Recommendation Subgraphs for Web Discovery [article]

Arda Antikacioglu, R. Ravi, Srinath Srihdar
2014 arXiv   pre-print
Our results confirm that it is not always necessary to implement complicated algorithms in the real-world and that very good practical results can be obtained by using heuristics that are backed by the  ...  In practice, solving these matching problems requires superlinear time and is not scalable.  ...  Indeed, when a = 1 or a = 2, its performance is comparable or better than greedy, though the difference is not as pronounced as it is in the simulations.  ... 
arXiv:1409.2042v1 fatcat:43duhly6w5cebdffoot5rp2pz4
« Previous Showing results 1 — 15 out of 22,799 results