Nurullah Sevim, Ege Ozan Özyedek, Furkan Şahinuç, Aykut Koç
Transformer-based language models utilize the attention mechanism for substantial performance improvements in almost all natural language processing (NLP) tasks. Similar attention structures are also extensively studied in several other areas. Although the attention mechanism enhances the model performances significantly, its quadratic complexity prevents efficient processing of long sequences. Recent works focused on eliminating the disadvantages of computational inefficiency and showed that transformer-based models can still reach competitive results without the attention layer. A pioneering study proposed the FNet, which replaces the attention layer with the Fourier Transform (FT) in the transformer encoder architecture. FNet achieves competitive performances concerning the original transformer encoder model while accelerating training process by removing the computational burden of the attention mechanism. However, the FNet model ignores essential properties of the FT from the classical signal processing that can be leveraged to increase model efficiency further. We propose different methods to deploy FT efficiently in transformer encoder models. Our proposed architectures have smaller number of model parameters, shorter training times, less memory usage, and some additional performance improvements. We demonstrate these improvements through extensive experiments on common benchmarks.
{"title":"Fast-FNet: Accelerating Transformer Encoder Models via Efficient Fourier Layers","authors":"Nurullah Sevim, Ege Ozan Özyedek, Furkan Şahinuç, Aykut Koç","doi":"arxiv-2209.12816","DOIUrl":"https://doi.org/arxiv-2209.12816","url":null,"abstract":"Transformer-based language models utilize the attention mechanism for\u0000substantial performance improvements in almost all natural language processing\u0000(NLP) tasks. Similar attention structures are also extensively studied in\u0000several other areas. Although the attention mechanism enhances the model\u0000performances significantly, its quadratic complexity prevents efficient\u0000processing of long sequences. Recent works focused on eliminating the\u0000disadvantages of computational inefficiency and showed that transformer-based\u0000models can still reach competitive results without the attention layer. A\u0000pioneering study proposed the FNet, which replaces the attention layer with the\u0000Fourier Transform (FT) in the transformer encoder architecture. FNet achieves\u0000competitive performances concerning the original transformer encoder model\u0000while accelerating training process by removing the computational burden of the\u0000attention mechanism. However, the FNet model ignores essential properties of\u0000the FT from the classical signal processing that can be leveraged to increase\u0000model efficiency further. We propose different methods to deploy FT efficiently\u0000in transformer encoder models. Our proposed architectures have smaller number\u0000of model parameters, shorter training times, less memory usage, and some\u0000additional performance improvements. We demonstrate these improvements through\u0000extensive experiments on common benchmarks.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544242","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rihab Gorsane, Omayma Mahjoub, Ruan de Kock, Roland Dubb, Siddarth Singh, Arnu Pretorius
Multi-agent reinforcement learning (MARL) has emerged as a useful approach to solving decentralised decision-making problems at scale. Research in the field has been growing steadily with many breakthrough algorithms proposed in recent years. In this work, we take a closer look at this rapid development with a focus on evaluation methodologies employed across a large body of research in cooperative MARL. By conducting a detailed meta-analysis of prior work, spanning 75 papers accepted for publication from 2016 to 2022, we bring to light worrying trends that put into question the true rate of progress. We further consider these trends in a wider context and take inspiration from single-agent RL literature on similar issues with recommendations that remain applicable to MARL. Combining these recommendations, with novel insights from our analysis, we propose a standardised performance evaluation protocol for cooperative MARL. We argue that such a standard protocol, if widely adopted, would greatly improve the validity and credibility of future research, make replication and reproducibility easier, as well as improve the ability of the field to accurately gauge the rate of progress over time by being able to make sound comparisons across different works. Finally, we release our meta-analysis data publicly on our project website for future research on evaluation: https://sites.google.com/view/marl-standard-protocol
{"title":"Towards a Standardised Performance Evaluation Protocol for Cooperative MARL","authors":"Rihab Gorsane, Omayma Mahjoub, Ruan de Kock, Roland Dubb, Siddarth Singh, Arnu Pretorius","doi":"arxiv-2209.10485","DOIUrl":"https://doi.org/arxiv-2209.10485","url":null,"abstract":"Multi-agent reinforcement learning (MARL) has emerged as a useful approach to\u0000solving decentralised decision-making problems at scale. Research in the field\u0000has been growing steadily with many breakthrough algorithms proposed in recent\u0000years. In this work, we take a closer look at this rapid development with a\u0000focus on evaluation methodologies employed across a large body of research in\u0000cooperative MARL. By conducting a detailed meta-analysis of prior work,\u0000spanning 75 papers accepted for publication from 2016 to 2022, we bring to\u0000light worrying trends that put into question the true rate of progress. We\u0000further consider these trends in a wider context and take inspiration from\u0000single-agent RL literature on similar issues with recommendations that remain\u0000applicable to MARL. Combining these recommendations, with novel insights from\u0000our analysis, we propose a standardised performance evaluation protocol for\u0000cooperative MARL. We argue that such a standard protocol, if widely adopted,\u0000would greatly improve the validity and credibility of future research, make\u0000replication and reproducibility easier, as well as improve the ability of the\u0000field to accurately gauge the rate of progress over time by being able to make\u0000sound comparisons across different works. Finally, we release our meta-analysis\u0000data publicly on our project website for future research on evaluation:\u0000https://sites.google.com/view/marl-standard-protocol","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"76 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Maja Schneider, Christian Marchington, Marco Körner
Expansive, informative datasets are vital in providing foundations and possibilities for scientific research and development across many fields of study. Assembly of grand datasets, however, frequently poses difficulty for the author and stakeholders alike, with a variety of considerations required throughout the collaboration efforts and development lifecycle. In this work, we discuss and analyse the challenges and opportunities we faced throughout the creation of a transnational, European agricultural dataset containing reference labels of cultivated crops. Together, this forms a succinct framework of important elements one should consider when forging a dataset of their own.
{"title":"Challenges and Opportunities of Large Transnational Datasets: A Case Study on European Administrative Crop Data","authors":"Maja Schneider, Christian Marchington, Marco Körner","doi":"arxiv-2210.07178","DOIUrl":"https://doi.org/arxiv-2210.07178","url":null,"abstract":"Expansive, informative datasets are vital in providing foundations and\u0000possibilities for scientific research and development across many fields of\u0000study. Assembly of grand datasets, however, frequently poses difficulty for the\u0000author and stakeholders alike, with a variety of considerations required\u0000throughout the collaboration efforts and development lifecycle. In this work,\u0000we discuss and analyse the challenges and opportunities we faced throughout the\u0000creation of a transnational, European agricultural dataset containing reference\u0000labels of cultivated crops. Together, this forms a succinct framework of\u0000important elements one should consider when forging a dataset of their own.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544236","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Phishing is a form of cybercrime and a threat that allows criminals, phishers, to deceive end users in order to steal their confidential and sensitive information. Attackers usually attempt to manipulate the psychology and emotions of victims. The increasing threat of phishing has made its study worthwhile and much research has been conducted into the issue. This paper explores the emotional factors that have been reported in previous studies to be significant in phishing victimization. In addition, we compare what security organizations and researchers have highlighted in terms of phishing types and categories as well as training in tackling the problem, in a literature review which takes into account all major credible and published sources.
{"title":"An Overview of Phishing Victimization: Human Factors, Training and the Role of Emotions","authors":"Mousa Jari","doi":"arxiv-2209.11197","DOIUrl":"https://doi.org/arxiv-2209.11197","url":null,"abstract":"Phishing is a form of cybercrime and a threat that allows criminals,\u0000phishers, to deceive end users in order to steal their confidential and\u0000sensitive information. Attackers usually attempt to manipulate the psychology\u0000and emotions of victims. The increasing threat of phishing has made its study\u0000worthwhile and much research has been conducted into the issue. This paper\u0000explores the emotional factors that have been reported in previous studies to\u0000be significant in phishing victimization. In addition, we compare what security\u0000organizations and researchers have highlighted in terms of phishing types and\u0000categories as well as training in tackling the problem, in a literature review\u0000which takes into account all major credible and published sources.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"45 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544230","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yanchao Xu, Wenbo Shao, Jun Li, Kai Yang, Weida Wang, Hua Huang, Chen Lv, Hong Wang
Intersection is one of the most challenging scenarios for autonomous driving tasks. Due to the complexity and stochasticity, essential applications (e.g., behavior modeling, motion prediction, safety validation, etc.) at intersections rely heavily on data-driven techniques. Thus, there is an intense demand for trajectory datasets of traffic participants (TPs) in intersections. Currently, most intersections in urban areas are equipped with traffic lights. However, there is not yet a large-scale, high-quality, publicly available trajectory dataset for signalized intersections. Therefore, in this paper, a typical two-phase signalized intersection is selected in Tianjin, China. Besides, a pipeline is designed to construct a Signalized INtersection Dataset (SIND), which contains 7 hours of recording including over 13,000 TPs with 7 types. Then, the behaviors of traffic light violations in SIND are recorded. Furthermore, the SIND is also compared with other similar works. The features of the SIND can be summarized as follows: 1) SIND provides more comprehensive information, including traffic light states, motion parameters, High Definition (HD) map, etc. 2) The category of TPs is diverse and characteristic, where the proportion of vulnerable road users (VRUs) is up to 62.6% 3) Multiple traffic light violations of non-motor vehicles are shown. We believe that SIND would be an effective supplement to existing datasets and can promote related research on autonomous driving.The dataset is available online via: https://github.com/SOTIF-AVLab/SinD
{"title":"SIND: A Drone Dataset at Signalized Intersection in China","authors":"Yanchao Xu, Wenbo Shao, Jun Li, Kai Yang, Weida Wang, Hua Huang, Chen Lv, Hong Wang","doi":"arxiv-2209.02297","DOIUrl":"https://doi.org/arxiv-2209.02297","url":null,"abstract":"Intersection is one of the most challenging scenarios for autonomous driving\u0000tasks. Due to the complexity and stochasticity, essential applications (e.g.,\u0000behavior modeling, motion prediction, safety validation, etc.) at intersections\u0000rely heavily on data-driven techniques. Thus, there is an intense demand for\u0000trajectory datasets of traffic participants (TPs) in intersections. Currently,\u0000most intersections in urban areas are equipped with traffic lights. However,\u0000there is not yet a large-scale, high-quality, publicly available trajectory\u0000dataset for signalized intersections. Therefore, in this paper, a typical\u0000two-phase signalized intersection is selected in Tianjin, China. Besides, a\u0000pipeline is designed to construct a Signalized INtersection Dataset (SIND),\u0000which contains 7 hours of recording including over 13,000 TPs with 7 types.\u0000Then, the behaviors of traffic light violations in SIND are recorded.\u0000Furthermore, the SIND is also compared with other similar works. The features\u0000of the SIND can be summarized as follows: 1) SIND provides more comprehensive\u0000information, including traffic light states, motion parameters, High Definition\u0000(HD) map, etc. 2) The category of TPs is diverse and characteristic, where the\u0000proportion of vulnerable road users (VRUs) is up to 62.6% 3) Multiple traffic\u0000light violations of non-motor vehicles are shown. We believe that SIND would be\u0000an effective supplement to existing datasets and can promote related research\u0000on autonomous driving.The dataset is available online via:\u0000https://github.com/SOTIF-AVLab/SinD","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"61 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544232","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The most pressing problems in science are neither empirical nor theoretical, but infrastructural. Scientific practice is defined by coproductive, mutually reinforcing infrastructural deficits and incentive systems that everywhere constrain and contort our art of curiosity in service of profit and prestige. Our infrastructural problems are not unique to science, but reflective of the broader logic of digital enclosure where platformatized control of information production and extraction fuels some of the largest corporations in the world. I have taken lessons learned from decades of intertwined digital cultures within and beyond academia like wikis, pirates, and librarians in order to draft a path towards more liberatory infrastructures for both science and society. Based on a system of peer-to-peer linked data, I sketch interoperable systems for shared data, tools, and knowledge that map onto three domains of platform capture: storage, computation and communication. The challenge of infrastructure is not solely technical, but also social and cultural, and so I attempt to ground a practical development blueprint in an ethics for organizing and maintaining it. I intend this draft as a rallying call for organization, to be revised with the input of collaborators and through the challenges posed by its implementation. I argue that a more liberatory future for science is neither utopian nor impractical -- the truly impractical choice is to continue to organize science as prestige fiefdoms resting on a pyramid scheme of underpaid labor, playing out the clock as every part of our work is swallowed whole by circling information conglomerates. It was arguably scientists looking for a better way to communicate that created something as radical as the internet in the first place, and I believe we can do it again.
{"title":"Decentralized Infrastructure for (Neuro)science","authors":"Jonny L. Saunders","doi":"arxiv-2209.07493","DOIUrl":"https://doi.org/arxiv-2209.07493","url":null,"abstract":"The most pressing problems in science are neither empirical nor theoretical,\u0000but infrastructural. Scientific practice is defined by coproductive, mutually\u0000reinforcing infrastructural deficits and incentive systems that everywhere\u0000constrain and contort our art of curiosity in service of profit and prestige.\u0000Our infrastructural problems are not unique to science, but reflective of the\u0000broader logic of digital enclosure where platformatized control of information\u0000production and extraction fuels some of the largest corporations in the world.\u0000I have taken lessons learned from decades of intertwined digital cultures\u0000within and beyond academia like wikis, pirates, and librarians in order to\u0000draft a path towards more liberatory infrastructures for both science and\u0000society. Based on a system of peer-to-peer linked data, I sketch interoperable\u0000systems for shared data, tools, and knowledge that map onto three domains of\u0000platform capture: storage, computation and communication. The challenge of\u0000infrastructure is not solely technical, but also social and cultural, and so I\u0000attempt to ground a practical development blueprint in an ethics for organizing\u0000and maintaining it. I intend this draft as a rallying call for organization, to\u0000be revised with the input of collaborators and through the challenges posed by\u0000its implementation. I argue that a more liberatory future for science is\u0000neither utopian nor impractical -- the truly impractical choice is to continue\u0000to organize science as prestige fiefdoms resting on a pyramid scheme of\u0000underpaid labor, playing out the clock as every part of our work is swallowed\u0000whole by circling information conglomerates. It was arguably scientists looking\u0000for a better way to communicate that created something as radical as the\u0000internet in the first place, and I believe we can do it again.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"210 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pengcheng He, Baolin Peng, Liyang Lu, Song Wang, Jie Mei, Yang Liu, Ruochen Xu, Hany Hassan Awadalla, Yu Shi, Chenguang Zhu, Wayne Xiong, Michael Zeng, Jianfeng Gao, Xuedong Huang
This paper presents Z-Code++, a new pre-trained language model optimized for abstractive text summarization. The model extends the state of the art encoder-decoder model using three techniques. First, we use a two-phase pre-training process to improve model's performance on low-resource summarization tasks. The model is first pre-trained using text corpora for language understanding, and then is continually pre-trained on summarization corpora for grounded text generation. Second, we replace self-attention layers in the encoder with disentangled attention layers, where each word is represented using two vectors that encode its content and position, respectively. Third, we use fusion-in-encoder, a simple yet effective method of encoding long sequences in a hierarchical manner. Z-Code++ creates new state of the art on 9 out of 13 text summarization tasks across 5 languages. Our model is parameter-efficient in that it outperforms the 600x larger PaLM-540B on XSum, and the finetuned 200x larger GPT3-175B on SAMSum. In zero-shot and few-shot settings, our model substantially outperforms the competing models.
{"title":"Z-Code++: A Pre-trained Language Model Optimized for Abstractive Summarization","authors":"Pengcheng He, Baolin Peng, Liyang Lu, Song Wang, Jie Mei, Yang Liu, Ruochen Xu, Hany Hassan Awadalla, Yu Shi, Chenguang Zhu, Wayne Xiong, Michael Zeng, Jianfeng Gao, Xuedong Huang","doi":"arxiv-2208.09770","DOIUrl":"https://doi.org/arxiv-2208.09770","url":null,"abstract":"This paper presents Z-Code++, a new pre-trained language model optimized for\u0000abstractive text summarization. The model extends the state of the art\u0000encoder-decoder model using three techniques. First, we use a two-phase\u0000pre-training process to improve model's performance on low-resource\u0000summarization tasks. The model is first pre-trained using text corpora for\u0000language understanding, and then is continually pre-trained on summarization\u0000corpora for grounded text generation. Second, we replace self-attention layers\u0000in the encoder with disentangled attention layers, where each word is\u0000represented using two vectors that encode its content and position,\u0000respectively. Third, we use fusion-in-encoder, a simple yet effective method of\u0000encoding long sequences in a hierarchical manner. Z-Code++ creates new state of\u0000the art on 9 out of 13 text summarization tasks across 5 languages. Our model\u0000is parameter-efficient in that it outperforms the 600x larger PaLM-540B on\u0000XSum, and the finetuned 200x larger GPT3-175B on SAMSum. In zero-shot and\u0000few-shot settings, our model substantially outperforms the competing models.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544301","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Early in the pandemic, we -- leaders in the research areas of programming languages (PL) and computer architecture (CA) -- realized that we had a problem: the only way to form new lasting connections in the community was to already have lasting connections in the community. Both of our academic communities had wonderful short-term mentoring programs to address this problem, but it was clear that we needed long-term mentoring programs. Those of us in CA approached this scientifically, making an evidence-backed case for community-wide long-term mentoring. In the meantime, one of us in PL had impulsively launched an unofficial long-term mentoring program, founded on chaos and spreadsheets. In January 2021, the latter grew to an official cross-institutional long-term mentoring program called SIGPLAN-M; in January 2022, the former grew to Computer Architecture Long-term Mentoring (CALM). The impacts have been strong: SIGPLAN-M reaches 328 mentees and 234 mentors across 41 countries, and mentees have described it as "life changing" and "a career saver." And while CALM is in its pilot phase -- with 13 mentors and 21 mentees across 7 countries -- it has received very positive feedback. The leaders of SIGPLAN-M and CALM shared our designs, impacts, and challenges along the way. Now, we wish to share those with you. We hope this will kick-start a larger long-term mentoring effort across all of computer science.
{"title":"Long-Term Mentoring for Computer Science Researchers","authors":"Emily Ruppel, Sihang Liu, Elba Garza, Sukyoung Ryu, Alexandra Silva, Talia Ringer","doi":"arxiv-2208.04738","DOIUrl":"https://doi.org/arxiv-2208.04738","url":null,"abstract":"Early in the pandemic, we -- leaders in the research areas of programming\u0000languages (PL) and computer architecture (CA) -- realized that we had a\u0000problem: the only way to form new lasting connections in the community was to\u0000already have lasting connections in the community. Both of our academic\u0000communities had wonderful short-term mentoring programs to address this\u0000problem, but it was clear that we needed long-term mentoring programs. Those of us in CA approached this scientifically, making an evidence-backed\u0000case for community-wide long-term mentoring. In the meantime, one of us in PL\u0000had impulsively launched an unofficial long-term mentoring program, founded on\u0000chaos and spreadsheets. In January 2021, the latter grew to an official\u0000cross-institutional long-term mentoring program called SIGPLAN-M; in January\u00002022, the former grew to Computer Architecture Long-term Mentoring (CALM). The impacts have been strong: SIGPLAN-M reaches 328 mentees and 234 mentors\u0000across 41 countries, and mentees have described it as \"life changing\" and \"a\u0000career saver.\" And while CALM is in its pilot phase -- with 13 mentors and 21\u0000mentees across 7 countries -- it has received very positive feedback. The\u0000leaders of SIGPLAN-M and CALM shared our designs, impacts, and challenges along\u0000the way. Now, we wish to share those with you. We hope this will kick-start a\u0000larger long-term mentoring effort across all of computer science.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-08-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544228","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The first two doctoral-level degrees in Computer Science in the US were awarded in June 1965. This paper discusses one of the degree recipients, Sister Mary Kenneth Keller, BVM.
{"title":"Mary Kenneth Keller: First US PhD in Computer Science","authors":"Jennifer Head, Dianne P. O'Leary","doi":"arxiv-2208.01765","DOIUrl":"https://doi.org/arxiv-2208.01765","url":null,"abstract":"The first two doctoral-level degrees in Computer Science in the US were\u0000awarded in June 1965. This paper discusses one of the degree recipients, Sister\u0000Mary Kenneth Keller, BVM.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"64 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544240","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Viktor Zobernig, Richard A. Saldanha, Jinke He, Erica van der Sar, Jasper van Doorn, Jia-Chen Hua, Lachlan R. Mason, Aleksander Czechowski, Drago Indjic, Tomasz Kosmala, Alessandro Zocca, Sandjai Bhulai, Jorge Montalvo Arvizu, Claude Klöckl, John Moriarty
The RangL project hosted by The Alan Turing Institute aims to encourage the wider uptake of reinforcement learning by supporting competitions relating to real-world dynamic decision problems. This article describes the reusable code repository developed by the RangL team and deployed for the 2022 Pathways to Net Zero Challenge, supported by the UK Net Zero Technology Centre. The winning solutions to this particular Challenge seek to optimize the UK's energy transition policy to net zero carbon emissions by 2050. The RangL repository includes an OpenAI Gym reinforcement learning environment and code that supports both submission to, and evaluation in, a remote instance of the open source EvalAI platform as well as all winning learning agent strategies. The repository is an illustrative example of RangL's capability to provide a reusable structure for future challenges.
{"title":"RangL: A Reinforcement Learning Competition Platform","authors":"Viktor Zobernig, Richard A. Saldanha, Jinke He, Erica van der Sar, Jasper van Doorn, Jia-Chen Hua, Lachlan R. Mason, Aleksander Czechowski, Drago Indjic, Tomasz Kosmala, Alessandro Zocca, Sandjai Bhulai, Jorge Montalvo Arvizu, Claude Klöckl, John Moriarty","doi":"arxiv-2208.00003","DOIUrl":"https://doi.org/arxiv-2208.00003","url":null,"abstract":"The RangL project hosted by The Alan Turing Institute aims to encourage the\u0000wider uptake of reinforcement learning by supporting competitions relating to\u0000real-world dynamic decision problems. This article describes the reusable code\u0000repository developed by the RangL team and deployed for the 2022 Pathways to\u0000Net Zero Challenge, supported by the UK Net Zero Technology Centre. The winning\u0000solutions to this particular Challenge seek to optimize the UK's energy\u0000transition policy to net zero carbon emissions by 2050. The RangL repository\u0000includes an OpenAI Gym reinforcement learning environment and code that\u0000supports both submission to, and evaluation in, a remote instance of the open\u0000source EvalAI platform as well as all winning learning agent strategies. The\u0000repository is an illustrative example of RangL's capability to provide a\u0000reusable structure for future challenges.","PeriodicalId":501533,"journal":{"name":"arXiv - CS - General Literature","volume":"193 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138544235","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：