Filters








1,111 Hits in 4.8 sec

Optimizing AD Pruning of Sponsored Search with Reinforcement Learning [article]

Yijiang Lian, Zhijie Chen, Xin Pei, Shuang Li, Yifei Wang, Yuefeng Qiu, Zhiheng Zhang, Zhipeng Tao, Liang Yuan, Hanju Guan, Kefeng Zhang, Zhigang Li (+1 others)
2020 arXiv   pre-print
A query with high commercial value might retrieve a great deal of ad candidates such that the ranking module could not afford.  ...  Industrial sponsored search system (SSS) can be logically divided into three modules: keywords matching, ad retrieving, and ranking. During ad retrieving, the ad candidates grow exponentially.  ...  Optimizing AD Pruning of Sponsored Search with Reinforcement Learning Conference'17, July 2017, Washington, DC, USA EXPERIMENTS 5.1 Setup Experiments are conducted on Baidu's sponsor search system.  ... 
arXiv:2008.02014v1 fatcat:k6jy4tsxczg6nicst6j5nduosa

SwiftPruner: Reinforced Evolutionary Pruning for Efficient Ad Relevance [article]

Li Lyna Zhang, Youkow Homma, Yujing Wang, Min Wu, Mao Yang, Ruofei Zhang, Ting Cao, Wei Shen
2022 arXiv   pre-print
Different from existing evolution algorithms that conduct random mutations, we propose a reinforced mutator with a latency-aware multi-objective reward to conduct better mutations for efficiently searching  ...  Our challenge is that previous methods typically prune all layers of the transformer to a high, uniform sparsity, thereby producing models which cannot achieve satisfactory inference speed with an acceptable  ...  INTRODUCTION In sponsored search, ad relevance measures the semantic similarity between a user's search query and an ad.  ... 
arXiv:2209.00625v1 fatcat:vga3ovpbingrlorqyyzkrbjwzu

Diversity driven Query Rewriting in Search Advertising [article]

Akash Kumar Mohankumar, Nikit Begwani, Amit Singh
2021 arXiv   pre-print
In this work, we introduce CLOVER, a framework to generate both high-quality and diverse rewrites by optimizing for human assessment of rewrite quality using our diversity-driven reinforcement learning  ...  For head and torso search queries, sponsored search engines use a huge repository of same intent queries and keywords, mined ahead of time.  ...  In sponsored search, advertisers bid on keywords relevant to their business to place their ads along with the organic search results.  ... 
arXiv:2106.03816v1 fatcat:2oak2rrn5za4hdktscajmh7ioi

Design of Sponsored Search Auction Mechanism for Federated Learning Advertising Platform

Hong Jiang, Tianxu Cui, Kaiwen Yang, Daqing Gong
2022 Computational Intelligence and Neuroscience  
A sponsored search auction mechanism design method is introduced to solve the problem of ranking the presentation order of participant advertisements.  ...  Due to the potential malicious bidding problem, which occurs when using the classic sponsored search auction mechanism under the federated learning scenario, this paper proposes a novel federated sponsored  ...  [46] have optimized the advertising pruning of sponsored search based on reinforcement learning. is is the first time that reinforcement learning technology has been used to address this problem.  ... 
doi:10.1155/2022/5787491 pmid:35432522 pmcid:PMC9010167 fatcat:v63bij76mjeq5eceglsvezpe3u

Machine Learning for Microcontroller-Class Hardware – A Review [article]

Swapnil Sayan Saha, Sandeep Singh Sandha, Mani Srivastava
2022 arXiv   pre-print
This paper highlights the unique requirements of enabling onboard machine learning for microcontroller class devices.  ...  The advancements in machine learning opened a new opportunity to bring intelligence to the low-end Internet-of-Things nodes such as microcontrollers.  ...  This is done using reinforcement learning (RL), one-shot gradient-driven NAS, evolutionary algorithms (with weight sharing), or Bayesian optimization [134] .  ... 
arXiv:2205.14550v3 fatcat:y272riitirhwfgfiotlwv5i7nu

Program Synthesis Using Deduction-Guided Reinforcement Learning [chapter]

Yanju Chen, Chenglong Wang, Osbert Bastani, Isil Dillig, Yu Feng
2020 Lecture Notes in Computer Science  
Second, it leverages deduction not only to prune the search space but also to guide search.  ...  In this paper, we present a new program synthesis algorithm based on reinforcement learning.  ...  Given an MDP M, the goal of reinforcement learning is to compute an optimal policy π * for M.  ... 
doi:10.1007/978-3-030-53291-8_30 fatcat:dxb5ws5fqvg6bkmsbba7aln35i

Exploring Lottery Ticket Hypothesis in Spiking Neural Networks [article]

Youngeun Kim, Yuhang Li, Hyoungseob Park, Yeshwanth Venkatesha, Ruokai Yin, Priyadarshini Panda
2022 arXiv   pre-print
However, the iterative searching process of LTH brings a huge training computational cost when combined with the multiple timesteps of SNNs.  ...  Most existing SNN pruning works focus on shallow SNNs (2~6 layers), however, deeper SNNs (>16 layers) are proposed by state-of-the-art SNN works, which is difficult to be compatible with the current SNN  ...  This work was supported in part by C-BRIC, a JUMP center sponsored by DARPA and SRC, Google Research Scholar Award, the National Science Foundation (Grant#1947826), TII (Abu Dhabi) and the DARPA AI Exploration  ... 
arXiv:2207.01382v2 fatcat:j7i7bgwlmfhpzo2vusdf7u4w74

Genetic Reinforcement Learning for Neurocontrol Problems [chapter]

Darrell Whitley, Stephen Dominic, Rajarshi Das, Charles W. Anderson
1993 Genetic Algorithms for Machine Learning  
On a simulated inverted-pendulum control problem, "genetic reinforcement learning" produces competitive results with AHC, another well-known reinforcement learning paradigm for neural networks that employs  ...  Empirical tests indicate that at least one class of genetic algorithms yields good performance for neural network weight optimization in terms of learning rates and scalability.  ...  CIAI is sponsored in part by the Colorado Advanced Technology Institute (CATI), an agency of the State of Colorado. Notes 1. The extra real value encodes the crossover probability of the string. 2.  ... 
doi:10.1007/978-1-4615-2740-4_5 fatcat:zzn26evgq5eddgzyoao3kciceu

Genetic reinforcement learning for neurocontrol problems

Darrell Whitley, Stephen Dominic, Rajarshi Das, Charles W. Anderson
1994 Machine Learning  
On a simulated inverted-pendulum control problem, "genetic reinforcement learning" produces competitive results with AHC, another well-known reinforcement learning paradigm for neural networks that employs  ...  Empirical tests indicate that at least one class of genetic algorithms yields good performance for neural network weight optimization in terms of learning rates and scalability.  ...  CIAI is sponsored in part by the Colorado Advanced Technology Institute (CATI), an agency of the State of Colorado. Notes 1. The extra real value encodes the crossover probability of the string. 2.  ... 
doi:10.1007/bf00993045 fatcat:m3ybyxp5u5bxpdml5nahpfnyfm

Hierarchically Constrained Adaptive Ad Exposure in Feeds [article]

Dagui Chen, Qi Yan, Chunjie Chen, Zhenzhe Zheng, Yangsu Liu, Zhenjia Ma and Chuan Yu, Jian Xu, Bo Zheng
2022 arXiv   pre-print
A contemporary feed application usually provides blended results of organic items and sponsored items~(ads) to users. Conventionally, ads are exposed at fixed positions.  ...  However, existing approaches to implementing the adaptive ad exposure still suffer from several limitations: 1) they usually fall into sub-optimal solutions because of only focusing on request-level optimization  ...  Learning-based methods [11, 23, 29, 30] mostly employ reinforcement learning (RL) [19] to search the optimal strategies and perform well in offline simulations.  ... 
arXiv:2205.15759v1 fatcat:cboltse35ze5dagfuzvjnovbxm

Maximum Margin Reward Networks for Learning from Explicit and Implicit Supervision

Haoruo Peng, Ming-Wei Chang, Wen-tau Yih
2017 Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing  
correctness of the predicted structure).  ...  However, annotated examples in structured domains are often costly to obtain, which thus limits the applications of neural networks.  ...  The first author is partly sponsored by DARPA under agreement number FA8750-13-2-0008. The U.S.  ... 
doi:10.18653/v1/d17-1252 dblp:conf/emnlp/PengCY17 fatcat:qs2ss7awszbqnc4h547vw2fghu

Model Learning for Look-Ahead Exploration in Continuous Control

Arpit Agarwal, Katharina Muelling, Katerina Fragkiadaki
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
We propose an exploration method that incorporates lookahead search over basic learnt skills and their dynamics, and use it for reinforcement learning (RL) of manipulation policies.  ...  We show that the proposed exploration strategy results in effective learning of complex manipulation policies faster than current state-of-the-art RL methods, and converges to better policies than methods  ...  We use learned skill dynamics with deep neural regressors and use them for look-ahead tree search, to guide effective exploration in reinforcement learning of complex manipulation tasks. path that leads  ... 
doi:10.1609/aaai.v33i01.33013151 fatcat:ddmhmwojjnf33glnplzpc5sbpu

Customized Monte Carlo Tree Search for LLVM/Polly's Composable Loop Optimization Transformations [article]

Jaehoon Koo, Prasanna Balaprakash, Michael Kruse, Xingfu Wu, Paul Hovland, Mary Hall
2021 arXiv   pre-print
Experimental results show that our MCTS algorithm finds pragma combinations with a speedup of 2.3x over Polly's heuristic optimizations on average.  ...  We compare our approach with random, greedy, and breadth-first search methods on PolyBench kernels and ECP proxy applications.  ...  Many approaches in the literature for solving loop optimization problems use ML and reinforcement learning (RL) and, recently, deep learning approaches.  ... 
arXiv:2105.04555v1 fatcat:jiduf3iel5fbfgmx5ni3ugapnq

Adversarial N-player Search using Locality for the Game of Battlesnake

Maximilian Benedikt Schier, Niclas Wüstenbecker
2019 Studierendenkonferenz Informatik  
Furthermore, the process of our heuristic parameter tuning with a grid search and a genetic algorithm is described.  ...  We propose to reduce the complexity of such games by limiting the search to players in the locality of the acting agent.  ...  Lastly, we want to thank all organizers and sponsors of the Battlesnake competition for making this event possible.  ... 
dblp:conf/skill/SchierW19 fatcat:2u2fylj3hrdrtodupd4n4zzy3a

Multi-Prize Lottery Ticket Hypothesis: Finding Accurate Binary Neural Networks by Pruning A Randomly Weighted Network [article]

James Diffenderfer, Bhavya Kailkhura
2021 arXiv   pre-print
) that (a) have comparable accuracy to a dense target network with learned weights (prize 1), (b) do not require any further training to achieve prize 1 (prize 2), and (c) is robust to extreme forms of  ...  However, finding these high performing trainable subnetworks is expensive, requiring iterative process of training and pruning weights.  ...  This document was prepared as an account of the work sponsored by an agency of the United States Government.  ... 
arXiv:2103.09377v1 fatcat:maay3daqvvegppeyah6aljhlyu
« Previous Showing results 1 — 15 out of 1,111 results