Ryan Marcus, assistant professor at the University of Pennsylvania. Using machine learning to build the next generation of data systems. 
      
    ____                       __  ___                          
   / __ \__  ______ _____     /  |/  /___ _____________  _______
  / /_/ / / / / __ `/ __ \   / /|_/ / __ `/ ___/ ___/ / / / ___/
 / _, _/ /_/ / /_/ / / / /  / /  / / /_/ / /  / /__/ /_/ (__  ) 
/_/ |_|\__, /\__,_/_/ /_/  /_/  /_/\__,_/_/   \___/\__,_/____/  
      /____/                                                    
        
   ___                   __  ___                    
  / _ \__ _____ ____    /  |/  /__ ___________ _____
 / , _/ // / _ `/ _ \  / /|_/ / _ `/ __/ __/ // (_-<
/_/|_|\_, /\_,_/_//_/ /_/  /_/\_,_/_/  \__/\_,_/___/
     /___/                                          
        
   ___  __  ___                    
  / _ \/  |/  /__ ___________ _____
 / , _/ /|_/ / _ `/ __/ __/ // (_-<
/_/|_/_/  /_/\_,_/_/  \__/\_,_/___/

Portfolio

I'm Ryan Marcus, an assistant professor at UPenn. I study applications of machine learning to systems, especially databases. I received my Ph.D. from Brandeis University, where I was advised by Olga Papaemmanouil. I was then a postdoc at MIT under the supervision of Tim Kraska.

Publications

  • Towards Full Stack Adaptivity in Permissioned Blockchains
    1. Chenyuan Wu
    2. Mohammad Javad Amiri
    3. Haoyun Qin
    4. Bhavana Mehta
    5. Ryan Marcus
    6. Boon Thau Loo
    VLDB '24 (pdf) (doi)
  • BFTGym: An Interactive Playground for BFT Protocols Demo.
    1. Haoyun Qin
    2. Chenyuan Wu
    3. Mohammad Javad Amiri
    4. Ryan Marcus
    5. Boon Thau Loo
    VLDB '24 (pdf) (doi)
  • Towards Truly Adaptive Byzantine Fault-Tolerant Consensus
    1. Chenyuan Wu
    2. Haoyun Qin
    3. Mohammad Javad Amiri
    4. Boon Thau Loo
    5. Dahlia Malkhi
    6. Ryan Marcus
    SIGOPS '24 (pdf) (doi)
  • Stage: Query Execution Time Prediction in Amazon Redshift
    1. Ziniu Wu
    2. Ryan Marcus
    3. Zhengchun Liu
    4. Parimarjan Negi
    5. Vikram Nathan
    6. Pascal Pfeil
    7. Gaurav Saxena
    8. Mohammad Rahman
    9. Balakrishnan Narayanaswamy
    10. Tim Kraska
    SIGMOD '24 (pdf) (doi)
  • Low Rank Approximation for Learned Query Optimization
    1. Zixuan Yi
    2. Yao Tian
    3. Zachary G. Ives
    4. Ryan Marcus
    aiDM @ SIGMOD '24 (pdf) (doi)
  • The Unreasonable Effectiveness of LLMs for Query Optimization
    1. Peter Akioyamen
    2. Zixuan Yi
    3. Ryan Marcus
    MLForSystems @ NeurIPS '24 (pdf) (doi)
  • AdaChain: A Learned Adaptive Blockchain
    1. Chenyuan Wu
    2. Bhavana Mehta
    3. Mohammad Javad Amiri
    4. Ryan Marcus
    5. Boon Thau Loo
    VLDB '23 (pdf) (doi)
  • AutoSteer: Learned Query Optimization for Any SQL Database
    1. Christoph Anneser
    2. Nesime Tatbul
    3. David Cohen
    4. Zhenggang Xu
    5. Prithvi Pandian
    6. Nikolay Leptev
    7. Ryan Marcus
    VLDB '23 (pdf) (doi)
  • Robust Query Driven Cardinality Estimation under Changing Workloads
    1. Parimarjan Negi
    2. Ziniu Wu
    3. Andreas Kipf
    4. Nesime Tatbul
    5. Ryan Marcus
    6. Sam Madden
    7. Tim Kraska
    8. Mohammad Alizadeh
    VLDB '23 (pdf) (doi)
  • QO-Insight: Inspecting Steered Query Optimizers Demo.
    1. Christoph Anneser
    2. Mario Petruccelli
    3. Nesime Tatbul
    4. David Cohen
    5. Zhenggang Xu
    6. Prithviraj Pandian
    7. Nikolay Laptev
    8. Ryan Marcus
    9. Alfons Kemper
    VLDB '23 (pdf) (doi)
  • Kepler: Robust Learning for Faster Parametric Query Optimization
    1. Lyric Doshi
    2. Vincent Zhuang
    3. Gaurav Jain
    4. Ryan Marcus
    5. Haoyu Huang
    6. Deniz Altinbuken
    7. Eugene Brevdo
    8. Campbell Fraser
    SIGMOD '23 (pdf) (doi)
  • Auto-WLM: Machine Learning Enhanced Workload Management in Amazon Redshift
    1. Gaurav Saxena
    2. Mohammad Rahman
    3. Naresh Chainani
    4. Chunbin Lin
    5. George Caragea
    6. Fahim Chowdhury
    7. Ryan Marcus
    8. Tim Kraska
    9. Ippokratis Pandis
    10. Balakrishnan (Murali) Narayanaswamy
    SIGMOD '23 (pdf) (doi)
  • Learned Query Superoptimization
    1. Ryan Marcus
    AIDB@VLDB '23 (pdf)
  • Towards Adaptive Fault-Tolerant Sharded Databases
    1. Bhavana Mehta
    2. Neelesh Chinnakonda Ashok Kumar
    3. Prashanth S Iyer
    4. Mohammad Javad Amiri
    5. Boon Thau Loo
    6. Ryan Marcus
    AIDB@VLDB '23 (pdf)
  • SageDB: An Instance-Optimized Data Analytics System
    1. Jialin Ding
    2. Ryan Marcus
    3. Andreas Kipf
    4. Vikram Nathan
    5. Aniruddha Nrusimha
    6. Kapil Vaidya
    7. Alexander van Renen
    8. Tim Kraska
    VLDB '22 (pdf) (doi)
  • Bao: Making Learned Query Optimization Practical
    1. Ryan Marcus
    2. Parimarjan Negi
    3. Hongzi Mao
    4. Nesime Tatbul
    5. Mohammad Alizadeh
    6. Tim Kraska
    SIGMOD Rec '22 (pdf) research highlight
  • LSI: A Learned Secondary Index Structure
    1. Andreas Kipf
    2. Dominik Horn
    3. Pascal Pfeil
    4. Ryan Marcus
    5. Tim Kraska
    aiDM @ SIGMOD '22 (pdf)
  • Benchmarking Learned Indexes
    1. Ryan Marcus
    2. Andreas Kipf
    3. Alexander Van Renen
    4. Mihail Stoian
    5. Sanchit Misra
    6. Alfons Kemper
    7. Thomas Neumann
    8. Tim Kraska
    VLDB '21 (pdf) (doi)
  • Flow-loss: Learning cardinality estimates that matter
    1. Parimarjan Negi
    2. Ryan Marcus
    3. Andreas Kipf
    4. Hongzi Mao
    5. Nesime Tatbul
    6. Tim Kraska
    7. Mohammad Alizadeh
    VLDB '21 (pdf) (doi)
  • Bao: Making Learned Query Optimization Practical
    1. Ryan Marcus
    2. Parimarjan Negi
    3. Hongzi Mao
    4. Nesime Tatbul
    5. Mohammad Alizadeh
    6. Tim Kraska
    SIGMOD '21 (pdf) (doi) best paper award
  • Steering Query Optimizers: A Practical Take on Big Data Workloads
    1. Parimarjan Negi
    2. Matteo Interlandi
    3. Ryan Marcus
    4. Mohammad Alizadeh
    5. Tim Kraska
    6. Marc Friedman
    7. Alekh Jindal
    SIGMOD '21 (pdf) (doi) best paper honorable mention
  • LEA: A Learned Encoding Advisor for Column Stores
    1. Lujing Cen
    2. Andreas Kipf
    3. Ryan Marcus
    4. Tim Kraska
    aiDM @ SIGMOD '21 (pdf) (doi)
  • Towards Practical Learned Indexing
    1. Mihail Stoian
    2. Andreas Kipf
    3. Ryan Marcus
    4. Tim Kraska
    aiDB @ VLDB '21 (pdf)
  • Towards a Benchmark for Learned Systems
    1. Laurent Bindschaedler
    2. Andreas Kipf
    3. Tim Kraska
    4. Ryan Marcus
    5. Umar Farooq Minhas
    SMDB @ ICDE '21 (pdf) (doi)
  • ARDA: Automatic Relational Data Augmentation for Machine Learning
    1. Nadiia Chepurko
    2. Ryan Marcus
    3. Emanuel Zgraggen
    4. Raul Castro Fernandez
    5. Tim Kraska
    6. David Karger
    VLDB '20 (pdf) (doi)
  • CDFShop: Exploring and Optimizing Learned Index Structures Demo.
    1. Ryan Marcus
    2. Emily Zhang
    3. Tim Kraska
    SIGMOD '20 (pdf) (doi)
  • Low Bitrate Compression of Video with Dynamic Backgrounds
    1. Solomon Garber
    2. Ryan Marcus
    3. Antonella DiLillo
    4. James Storer
    DCC '20 (pdf) (doi)
  • RadixSpline: a single-pass learned index
    1. Andreas Kipf
    2. Ryan Marcus
    3. Alexander van Renen
    4. Mihail Stoian
    5. Alfons Kemper
    6. Tim Kraska
    7. Thomas Neumann
    aiDM @ SIGMOD '20 (pdf) (doi)
  • Cost-Guided Cardinality Estimation: Focus Where it Matters
    1. Parimarjan Negi
    2. Ryan Marcus
    3. Hongzi Mao
    4. Nesime Tatbul
    5. Tim Kraska
    6. Mohammad Alizadeh
    SMDB @ ICDE '20 (pdf)
  • Learned Garbage Collection
    1. Lujing Cen
    2. Ryan Marcus
    3. Hongzi Mao
    4. Justin Gottschlich
    5. Mohammad Alizadeh
    6. Tim Kraska
    MAPL @ PLDI '20 (pdf) (doi)
  • Buffer Pool Aware Query Scheduling via Deep Reinforcement Learning
    1. Chi Zhang
    2. Ryan Marcus
    3. Anat Kleiman
    4. Olga Papaemmanouil
    5. Bingsheng He
    6. Berthold Reinwald
    7. Yingjun Wu
    AIDB@VLDB '20 (pdf)
  • Plan-Structured Deep Neural Network Models for Query Performance Prediction
    1. Ryan Marcus
    2. Olga Papaemmanouil
    VLDB '19 (pdf) (doi)
  • Neo: A Learned Query Optimizer
    1. Ryan Marcus
    2. Parimarjan Negi
    3. Hongzi Mao
    4. Chi Zhang
    5. Mohammad Alizadeh
    6. Tim Kraska
    7. Olga Papaemmanouil
    8. Nesime Tatbul
    VLDB '19 (pdf) (doi)
  • NashDB: Fragmentation, Replication, and Provisioning using Economic Methods Demo.
    1. Ryan Marcus
    2. Chi Zhang
    3. Shuai Yu
    4. Geoffrey Kao
    5. Olga Papaemmanouil
    VLDB '19 (pdf) (doi)
  • AI Meets AI: Leveraging Query Executions to Improve Index Recommendations
    1. Bailu Ding
    2. Sudipto Das
    3. Ryan Marcus
    4. Wentao Wu
    5. Surajit Chaudhuri
    6. Vivek R. Narasayya
    SIGMOD '19 (pdf) (doi)
  • Park: An Open Platform for Learning-Augmented Computer Systems
    1. Hongzi Mao
    2. Parimarjan Negi
    3. Akshay Narayan
    4. Hanrui Wang
    5. Jiacheng Yang
    6. Haonan Wang
    7. Ryan Marcus
    8. ravichandra addanki
    9. Mehrdad Khani Shirkoohi
    10. Songtao He
    11. Vikram Nathan
    12. Frank Cangialosi
    13. Shaileshh Venkatakrishnan
    14. Wei-Hung Weng
    15. Song Han
    16. Tim Kraska
    17. Mohammad Alizadeh
    18. H. Wallach
    19. H. Larochelle
    20. A. Beygelzimer
    21. F. d Alche-Buc
    22. E. Fox
    23. R. Garnett
    NeurIPS '19 (pdf) (doi)
  • Compact Representations of Dynamic Video Background Using Motion Sprites
    1. Solomon Garber
    2. Aaditya Prakash
    3. Ryan Marcus
    4. Antonella DiLillo
    5. James Storer
    DCC '19 (pdf) (doi)
  • Towards a Hands-Free Query Optimizer through Deep Learning
    1. Ryan Marcus
    2. Olga Papaemmanouil
    CIDR '19 (pdf)
  • Park: An Open Platform for Learning-Augmented Computer Systems
    1. Hongzi Mao
    2. Parimarjan Negi
    3. Akshay Narayan
    4. Hanrui Wang
    5. Jiacheng Yang
    6. Haonan Wang
    7. Ryan Marcus
    8. Ravichandra Addanki
    9. Mehrdad Khani
    10. Songtao He
    11. Vikram Nathan
    12. Frank Cangialosi
    13. Shaileshh Bojja Venkatakrishnan
    14. Wei-Hung Weng
    15. Song Han
    16. Tim Kraska
    17. Mohammad Alizadeh
    RL4RealLife @ ICML '19 (pdf) best workshop paper award
  • SOSD: A Benchmark for Learned Indexes
    1. Andreas Kipf
    2. Ryan Marcus
    3. Alexander van Renen
    4. Mihail Stoian
    5. Alfons Kemper
    6. Tim Kraska
    7. Thomas Neumann
    MLForSystems @ NeurIPS '19 (pdf)
  • NashDB: An End-to-End Economic Method for Elastic Database Fragmentation, Replication, and Provisioning
    1. Ryan Marcus
    2. Olga Papaemmanouil
    3. Sofiya Semenova
    4. Solomon Garber
    SIGMOD '18 (pdf) (doi)
  • Deep Reinforcement Learning for Join Order Enumeration
    1. Ryan Marcus
    2. Olga Papaemmanouil
    aiDM @ SIGMOD '18 (pdf)
  • A Learning-Based Service for Cost and Performance Management of Cloud Databases Demo.
    1. Ryan Marcus
    2. Sofiya Semenova
    3. Olga Papaemmanouil
    ICDE '17 (pdf) (doi)
  • Releasing Cloud Databases from the Chains of Performance Prediction Models
    1. Ryan Marcus
    2. Olga Papaemmanouil
    CIDR '17 (pdf)
  • WiSeDB: A Learning-based Workload Management Advisor for Cloud Databases
    1. Ryan Marcus
    2. Olga Papaemmanouil
    VLDB '16 (pdf) (doi)
  • Workload Management for Cloud Databases via Machine Learning
    1. Ryan Marcus
    2. Olga Papaemmanouil
    CloudDM @ ICDE '16 (pdf) (doi)
  • Techniques for Automated Performance Analysis
    1. Ryan Marcus
    tech. report '14 (pdf)
  • An Efficient Algorithm and Monte Carlo Methods for Inferring Functional Dependencies
    1. Ryan Marcus
    2. Shaughan Lavine
    tech. report '14 (pdf)
  • DP: a Fast Median Filter Approximation
    1. Ryan Marcus
    2. William C. Ward
    tech. report '13 (pdf)
  • MCMini: Monte Carlo on GPGPU
    1. Ryan Marcus
    tech. report '12 (pdf)
  • Developing a Monte Carlo mini-app for exascale co-design
    1. Lawrence J. Cox
    2. Ryan Marcus
    tech. report '11 (pdf)

Become a Conservancy Supporter!

Copyright 2024 Ryan Marcus