--- _id: '6031' abstract: - lang: eng text: We introduce Clover, a new library for efficient computation using low-precision data, providing mathematical routines required by fundamental methods in optimization and sparse recovery. Our library faithfully implements variants of stochastic quantization that guarantee convergence at low precision, and supports data formats from 4-bit quantized to 32-bit IEEE-754 on current Intel processors. In particular, we show that 4-bit can be implemented efficiently using Intel AVX despite the lack of native support for this data format. Experimental results with dot product, matrix-vector multiplication (MVM), gradient descent (GD), and iterative hard thresholding (IHT) demonstrate that the attainable speedups are in many cases close to linear with respect to the reduction of precision due to reduced data movement. Finally, for GD and IHT, we show examples of absolute speedup achieved by 4-bit versus 32-bit, by iterating until a given target error is achieved. article_number: '8598402' article_processing_charge: No author: - first_name: Alen full_name: Stojanov, Alen last_name: Stojanov - first_name: Tyler Michael full_name: Smith, Tyler Michael last_name: Smith - first_name: Dan-Adrian full_name: Alistarh, Dan-Adrian id: 4A899BFC-F248-11E8-B48F-1D18A9856A87 last_name: Alistarh orcid: 0000-0003-3650-940X - first_name: Markus full_name: Puschel, Markus last_name: Puschel citation: ama: 'Stojanov A, Smith TM, Alistarh D-A, Puschel M. Fast quantized arithmetic on x86: Trading compute for data movement. In: 2018 IEEE International Workshop on Signal Processing Systems. Vol 2018-October. IEEE; 2018. doi:10.1109/SiPS.2018.8598402' apa: 'Stojanov, A., Smith, T. M., Alistarh, D.-A., & Puschel, M. (2018). Fast quantized arithmetic on x86: Trading compute for data movement. In 2018 IEEE International Workshop on Signal Processing Systems (Vol. 2018–October). Cape Town, South Africa: IEEE. https://doi.org/10.1109/SiPS.2018.8598402' chicago: 'Stojanov, Alen, Tyler Michael Smith, Dan-Adrian Alistarh, and Markus Puschel. “Fast Quantized Arithmetic on X86: Trading Compute for Data Movement.” In 2018 IEEE International Workshop on Signal Processing Systems, Vol. 2018–October. IEEE, 2018. https://doi.org/10.1109/SiPS.2018.8598402.' ieee: 'A. Stojanov, T. M. Smith, D.-A. Alistarh, and M. Puschel, “Fast quantized arithmetic on x86: Trading compute for data movement,” in 2018 IEEE International Workshop on Signal Processing Systems, Cape Town, South Africa, 2018, vol. 2018–October.' ista: 'Stojanov A, Smith TM, Alistarh D-A, Puschel M. 2018. Fast quantized arithmetic on x86: Trading compute for data movement. 2018 IEEE International Workshop on Signal Processing Systems. SiPS: Workshop on Signal Processing Systems vol. 2018–October, 8598402.' mla: 'Stojanov, Alen, et al. “Fast Quantized Arithmetic on X86: Trading Compute for Data Movement.” 2018 IEEE International Workshop on Signal Processing Systems, vol. 2018–October, 8598402, IEEE, 2018, doi:10.1109/SiPS.2018.8598402.' short: A. Stojanov, T.M. Smith, D.-A. Alistarh, M. Puschel, in:, 2018 IEEE International Workshop on Signal Processing Systems, IEEE, 2018. conference: end_date: 2018-10-24 location: Cape Town, South Africa name: 'SiPS: Workshop on Signal Processing Systems' start_date: 2018-10-21 date_created: 2019-02-17T22:59:25Z date_published: 2018-12-31T00:00:00Z date_updated: 2023-09-19T14:41:51Z day: '31' department: - _id: DaAl doi: 10.1109/SiPS.2018.8598402 external_id: isi: - '000465106800060' isi: 1 language: - iso: eng month: '12' oa_version: None publication: 2018 IEEE International Workshop on Signal Processing Systems publication_status: published publisher: IEEE quality_controlled: '1' scopus_import: '1' status: public title: 'Fast quantized arithmetic on x86: Trading compute for data movement' type: conference user_id: c635000d-4b10-11ee-a964-aac5a93f6ac1 volume: 2018-October year: '2018' ... --- _id: '25' abstract: - lang: eng text: 'Partially observable Markov decision processes (POMDPs) are the standard models for planning under uncertainty with both finite and infinite horizon. Besides the well-known discounted-sum objective, indefinite-horizon objective (aka Goal-POMDPs) is another classical objective for POMDPs. In this case, given a set of target states and a positive cost for each transition, the optimization objective is to minimize the expected total cost until a target state is reached. In the literature, RTDP-Bel or heuristic search value iteration (HSVI) have been used for solving Goal-POMDPs. Neither of these algorithms has theoretical convergence guarantees, and HSVI may even fail to terminate its trials. We give the following contributions: (1) We discuss the challenges introduced in Goal-POMDPs and illustrate how they prevent the original HSVI from converging. (2) We present a novel algorithm inspired by HSVI, termed Goal-HSVI, and show that our algorithm has convergence guarantees. (3) We show that Goal-HSVI outperforms RTDP-Bel on a set of well-known examples.' acknowledgement: '∗This work has been supported by Vienna Science and Technology Fund (WWTF) Project ICT15-003, Austrian Science Fund (FWF) NFN Grant No S11407-N23 (RiSE/SHiNE), and ERC Starting grant (279307: Graph Games). This research was sponsored by the Army Research Laboratory and was accomplished under Cooperative Agreement Number W911NF-13-2-0045 (ARL Cyber Security CRA). ' article_processing_charge: No author: - first_name: Karel full_name: Horák, Karel last_name: Horák - first_name: Branislav full_name: Bošanský, Branislav last_name: Bošanský - first_name: Krishnendu full_name: Chatterjee, Krishnendu id: 2E5DCA20-F248-11E8-B48F-1D18A9856A87 last_name: Chatterjee orcid: 0000-0002-4561-241X citation: ama: 'Horák K, Bošanský B, Chatterjee K. Goal-HSVI: Heuristic search value iteration for goal-POMDPs. In: Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence. Vol 2018-July. IJCAI; 2018:4764-4770. doi:10.24963/ijcai.2018/662' apa: 'Horák, K., Bošanský, B., & Chatterjee, K. (2018). Goal-HSVI: Heuristic search value iteration for goal-POMDPs. In Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (Vol. 2018–July, pp. 4764–4770). Stockholm, Sweden: IJCAI. https://doi.org/10.24963/ijcai.2018/662' chicago: 'Horák, Karel, Branislav Bošanský, and Krishnendu Chatterjee. “Goal-HSVI: Heuristic Search Value Iteration for Goal-POMDPs.” In Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, 2018–July:4764–70. IJCAI, 2018. https://doi.org/10.24963/ijcai.2018/662.' ieee: 'K. Horák, B. Bošanský, and K. Chatterjee, “Goal-HSVI: Heuristic search value iteration for goal-POMDPs,” in Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, Stockholm, Sweden, 2018, vol. 2018–July, pp. 4764–4770.' ista: 'Horák K, Bošanský B, Chatterjee K. 2018. Goal-HSVI: Heuristic search value iteration for goal-POMDPs. Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence. IJCAI: International Joint Conference on Artificial Intelligence vol. 2018–July, 4764–4770.' mla: 'Horák, Karel, et al. “Goal-HSVI: Heuristic Search Value Iteration for Goal-POMDPs.” Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, vol. 2018–July, IJCAI, 2018, pp. 4764–70, doi:10.24963/ijcai.2018/662.' short: K. Horák, B. Bošanský, K. Chatterjee, in:, Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, IJCAI, 2018, pp. 4764–4770. conference: end_date: 2018-07-19 location: Stockholm, Sweden name: 'IJCAI: International Joint Conference on Artificial Intelligence' start_date: 2018-07-13 date_created: 2018-12-11T11:44:13Z date_published: 2018-07-01T00:00:00Z date_updated: 2023-09-19T14:44:59Z day: '01' department: - _id: KrCh doi: 10.24963/ijcai.2018/662 ec_funded: 1 external_id: isi: - '000764175404127' isi: 1 language: - iso: eng main_file_link: - open_access: '1' url: https://doi.org/10.24963/ijcai.2018/662 month: '07' oa: 1 oa_version: Published Version page: 4764 - 4770 project: - _id: 25892FC0-B435-11E9-9278-68D0E5697425 grant_number: ICT15-003 name: Efficient Algorithms for Computer Aided Verification - _id: 25832EC2-B435-11E9-9278-68D0E5697425 call_identifier: FWF grant_number: S 11407_N23 name: Rigorous Systems Engineering - _id: 2581B60A-B435-11E9-9278-68D0E5697425 call_identifier: FP7 grant_number: '279307' name: 'Quantitative Graph Games: Theory and Applications' publication: Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence publication_status: published publisher: IJCAI publist_id: '8030' quality_controlled: '1' scopus_import: '1' status: public title: 'Goal-HSVI: Heuristic search value iteration for goal-POMDPs' type: conference user_id: c635000d-4b10-11ee-a964-aac5a93f6ac1 volume: 2018-July year: '2018' ... --- _id: '24' abstract: - lang: eng text: Partially-observable Markov decision processes (POMDPs) with discounted-sum payoff are a standard framework to model a wide range of problems related to decision making under uncertainty. Traditionally, the goal has been to obtain policies that optimize the expectation of the discounted-sum payoff. A key drawback of the expectation measure is that even low probability events with extreme payoff can significantly affect the expectation, and thus the obtained policies are not necessarily risk-averse. An alternate approach is to optimize the probability that the payoff is above a certain threshold, which allows obtaining risk-averse policies, but ignores optimization of the expectation. We consider the expectation optimization with probabilistic guarantee (EOPG) problem, where the goal is to optimize the expectation ensuring that the payoff is above a given threshold with at least a specified probability. We present several results on the EOPG problem, including the first algorithm to solve it. acknowledgement: "This research was supported by the Vienna Science and Technology Fund (WWTF) grant ICT15-003; Austrian Science Fund (FWF): S11407-N23(RiSE/SHiNE);and an ERC Start Grant (279307:Graph Games).\r\n" article_processing_charge: No author: - first_name: Krishnendu full_name: Chatterjee, Krishnendu id: 2E5DCA20-F248-11E8-B48F-1D18A9856A87 last_name: Chatterjee orcid: 0000-0002-4561-241X - first_name: Adrian full_name: Elgyütt, Adrian id: 4A2E9DBA-F248-11E8-B48F-1D18A9856A87 last_name: Elgyütt - first_name: Petr full_name: Novotny, Petr id: 3CC3B868-F248-11E8-B48F-1D18A9856A87 last_name: Novotny - first_name: Owen full_name: Rouillé, Owen last_name: Rouillé citation: ama: 'Chatterjee K, Elgyütt A, Novotný P, Rouillé O. Expectation optimization with probabilistic guarantees in POMDPs with discounted-sum objectives. In: Vol 2018. IJCAI; 2018:4692-4699. doi:10.24963/ijcai.2018/652' apa: 'Chatterjee, K., Elgyütt, A., Novotný, P., & Rouillé, O. (2018). Expectation optimization with probabilistic guarantees in POMDPs with discounted-sum objectives (Vol. 2018, pp. 4692–4699). Presented at the IJCAI: International Joint Conference on Artificial Intelligence, Stockholm, Sweden: IJCAI. https://doi.org/10.24963/ijcai.2018/652' chicago: Chatterjee, Krishnendu, Adrian Elgyütt, Petr Novotný, and Owen Rouillé. “Expectation Optimization with Probabilistic Guarantees in POMDPs with Discounted-Sum Objectives,” 2018:4692–99. IJCAI, 2018. https://doi.org/10.24963/ijcai.2018/652. ieee: 'K. Chatterjee, A. Elgyütt, P. Novotný, and O. Rouillé, “Expectation optimization with probabilistic guarantees in POMDPs with discounted-sum objectives,” presented at the IJCAI: International Joint Conference on Artificial Intelligence, Stockholm, Sweden, 2018, vol. 2018, pp. 4692–4699.' ista: 'Chatterjee K, Elgyütt A, Novotný P, Rouillé O. 2018. Expectation optimization with probabilistic guarantees in POMDPs with discounted-sum objectives. IJCAI: International Joint Conference on Artificial Intelligence vol. 2018, 4692–4699.' mla: Chatterjee, Krishnendu, et al. Expectation Optimization with Probabilistic Guarantees in POMDPs with Discounted-Sum Objectives. Vol. 2018, IJCAI, 2018, pp. 4692–99, doi:10.24963/ijcai.2018/652. short: K. Chatterjee, A. Elgyütt, P. Novotný, O. Rouillé, in:, IJCAI, 2018, pp. 4692–4699. conference: end_date: 2018-07-19 location: Stockholm, Sweden name: 'IJCAI: International Joint Conference on Artificial Intelligence' start_date: 2018-07-13 date_created: 2018-12-11T11:44:13Z date_published: 2018-07-01T00:00:00Z date_updated: 2023-09-19T14:45:48Z day: '01' department: - _id: KrCh - _id: ToHe doi: 10.24963/ijcai.2018/652 ec_funded: 1 external_id: arxiv: - '1804.10601' isi: - '000764175404117' intvolume: ' 2018' isi: 1 language: - iso: eng main_file_link: - open_access: '1' url: https://arxiv.org/abs/1804.10601 month: '07' oa: 1 oa_version: Preprint page: 4692 - 4699 project: - _id: 25892FC0-B435-11E9-9278-68D0E5697425 grant_number: ICT15-003 name: Efficient Algorithms for Computer Aided Verification - _id: 25832EC2-B435-11E9-9278-68D0E5697425 call_identifier: FWF grant_number: S 11407_N23 name: Rigorous Systems Engineering - _id: 2581B60A-B435-11E9-9278-68D0E5697425 call_identifier: FP7 grant_number: '279307' name: 'Quantitative Graph Games: Theory and Applications' publication_status: published publisher: IJCAI publist_id: '8031' quality_controlled: '1' scopus_import: '1' status: public title: Expectation optimization with probabilistic guarantees in POMDPs with discounted-sum objectives type: conference user_id: c635000d-4b10-11ee-a964-aac5a93f6ac1 volume: 2018 year: '2018' ... --- _id: '34' abstract: - lang: eng text: Partially observable Markov decision processes (POMDPs) are widely used in probabilistic planning problems in which an agent interacts with an environment using noisy and imprecise sensors. We study a setting in which the sensors are only partially defined and the goal is to synthesize “weakest” additional sensors, such that in the resulting POMDP, there is a small-memory policy for the agent that almost-surely (with probability 1) satisfies a reachability objective. We show that the problem is NP-complete, and present a symbolic algorithm by encoding the problem into SAT instances. We illustrate trade-offs between the amount of memory of the policy and the number of additional sensors on a simple example. We have implemented our approach and consider three classical POMDP examples from the literature, and show that in all the examples the number of sensors can be significantly decreased (as compared to the existing solutions in the literature) without increasing the complexity of the policies. alternative_title: - ICAPS article_processing_charge: No author: - first_name: Krishnendu full_name: Chatterjee, Krishnendu id: 2E5DCA20-F248-11E8-B48F-1D18A9856A87 last_name: Chatterjee orcid: 0000-0002-4561-241X - first_name: Martin full_name: Chemlík, Martin last_name: Chemlík - first_name: Ufuk full_name: Topcu, Ufuk last_name: Topcu citation: ama: 'Chatterjee K, Chemlík M, Topcu U. Sensor synthesis for POMDPs with reachability objectives. In: Vol 2018. AAAI Press; 2018:47-55.' apa: 'Chatterjee, K., Chemlík, M., & Topcu, U. (2018). Sensor synthesis for POMDPs with reachability objectives (Vol. 2018, pp. 47–55). Presented at the ICAPS: International Conference on Automated Planning and Scheduling, Delft, Netherlands: AAAI Press.' chicago: Chatterjee, Krishnendu, Martin Chemlík, and Ufuk Topcu. “Sensor Synthesis for POMDPs with Reachability Objectives,” 2018:47–55. AAAI Press, 2018. ieee: 'K. Chatterjee, M. Chemlík, and U. Topcu, “Sensor synthesis for POMDPs with reachability objectives,” presented at the ICAPS: International Conference on Automated Planning and Scheduling, Delft, Netherlands, 2018, vol. 2018, pp. 47–55.' ista: 'Chatterjee K, Chemlík M, Topcu U. 2018. Sensor synthesis for POMDPs with reachability objectives. ICAPS: International Conference on Automated Planning and Scheduling, ICAPS, vol. 2018, 47–55.' mla: Chatterjee, Krishnendu, et al. Sensor Synthesis for POMDPs with Reachability Objectives. Vol. 2018, AAAI Press, 2018, pp. 47–55. short: K. Chatterjee, M. Chemlík, U. Topcu, in:, AAAI Press, 2018, pp. 47–55. conference: end_date: 2018-06-29 location: Delft, Netherlands name: 'ICAPS: International Conference on Automated Planning and Scheduling' start_date: 2018-06-24 date_created: 2018-12-11T11:44:16Z date_published: 2018-06-01T00:00:00Z date_updated: 2023-09-19T14:44:14Z day: '01' department: - _id: KrCh ec_funded: 1 external_id: arxiv: - '1710.00675' isi: - '000492986200006' intvolume: ' 2018' isi: 1 language: - iso: eng main_file_link: - open_access: '1' url: https://arxiv.org/abs/1710.00675 month: '06' oa: 1 oa_version: Preprint page: 47 - 55 project: - _id: 2584A770-B435-11E9-9278-68D0E5697425 call_identifier: FWF grant_number: P 23499-N23 name: Modern Graph Algorithmic Techniques in Formal Verification - _id: 2581B60A-B435-11E9-9278-68D0E5697425 call_identifier: FP7 grant_number: '279307' name: 'Quantitative Graph Games: Theory and Applications' - _id: 25832EC2-B435-11E9-9278-68D0E5697425 call_identifier: FWF grant_number: S 11407_N23 name: Rigorous Systems Engineering - _id: 2587B514-B435-11E9-9278-68D0E5697425 name: Microsoft Research Faculty Fellowship publication_status: published publisher: AAAI Press publist_id: '8021' quality_controlled: '1' scopus_import: '1' status: public title: Sensor synthesis for POMDPs with reachability objectives type: conference user_id: c635000d-4b10-11ee-a964-aac5a93f6ac1 volume: 2018 year: '2018' ... --- _id: '18' abstract: - lang: eng text: An N-superconcentrator is a directed, acyclic graph with N input nodes and N output nodes such that every subset of the inputs and every subset of the outputs of same cardinality can be connected by node-disjoint paths. It is known that linear-size and bounded-degree superconcentrators exist. We prove the existence of such superconcentrators with asymptotic density 25.3 (where the density is the number of edges divided by N). The previously best known densities were 28 [12] and 27.4136 [17]. article_processing_charge: No author: - first_name: Vladimir full_name: Kolmogorov, Vladimir id: 3D50B0BA-F248-11E8-B48F-1D18A9856A87 last_name: Kolmogorov - first_name: Michal full_name: Rolinek, Michal id: 3CB3BC06-F248-11E8-B48F-1D18A9856A87 last_name: Rolinek citation: ama: Kolmogorov V, Rolinek M. Superconcentrators of density 25.3. Ars Combinatoria. 2018;141(10):269-304. apa: Kolmogorov, V., & Rolinek, M. (2018). Superconcentrators of density 25.3. Ars Combinatoria. Charles Babbage Research Centre. chicago: Kolmogorov, Vladimir, and Michal Rolinek. “Superconcentrators of Density 25.3.” Ars Combinatoria. Charles Babbage Research Centre, 2018. ieee: V. Kolmogorov and M. Rolinek, “Superconcentrators of density 25.3,” Ars Combinatoria, vol. 141, no. 10. Charles Babbage Research Centre, pp. 269–304, 2018. ista: Kolmogorov V, Rolinek M. 2018. Superconcentrators of density 25.3. Ars Combinatoria. 141(10), 269–304. mla: Kolmogorov, Vladimir, and Michal Rolinek. “Superconcentrators of Density 25.3.” Ars Combinatoria, vol. 141, no. 10, Charles Babbage Research Centre, 2018, pp. 269–304. short: V. Kolmogorov, M. Rolinek, Ars Combinatoria 141 (2018) 269–304. date_created: 2018-12-11T11:44:11Z date_published: 2018-10-01T00:00:00Z date_updated: 2023-09-19T14:46:18Z day: '01' department: - _id: VlKo external_id: arxiv: - '1405.7828' isi: - '000446809500022' intvolume: ' 141' isi: 1 issue: '10' language: - iso: eng main_file_link: - open_access: '1' url: https://arxiv.org/abs/1405.7828 month: '10' oa: 1 oa_version: Preprint page: 269 - 304 publication: Ars Combinatoria publication_identifier: issn: - 0381-7032 publication_status: published publisher: Charles Babbage Research Centre publist_id: '8037' quality_controlled: '1' scopus_import: '1' status: public title: Superconcentrators of density 25.3 type: journal_article user_id: c635000d-4b10-11ee-a964-aac5a93f6ac1 volume: 141 year: '2018' ...