Enabling Reproducibility in Machine Learning MLTrain@RML (ICML 2018)
Enabling Reproducibility in Machine Learning MLTrain@RML
This workshop focuses on how to present papers from the coding perspective so that reproducibility and replication of results in the Machine Learning community becomes easier.
Papers from the Machine Learning community are supposed to be a valuable
asset. They can help to inform and inspire future research. They can be a useful educational tool for students. They are the driving force of innovation and differentiation in the industry, so quick and accurate implementation is really critical. On the research side they can help us answer the most fundamental questions about our existence – what does it mean to learn and what does it mean to be human? Reproducibility, while not always possible in science (consider the study of a transient astrological phenomenon like a passing comet), is a powerful criteria for improving the quality of research. A result which is reproducible is more likely to be robust and meaningful and rules out many types of experimenter error (either fraud or accidental). There are many interesting open questions about how reproducibility issues intersect with the Machine Learning community:
- How can we tell if papers in the Machine Learning community are reproducible even in theory? If a paper is about recommending news sites before a particular election, and the results come from running the system online in production – it will be impossible to reproduce the published results because the state of the world is irreversibly changed from when the experiment was run.
- What does it mean for a paper to be reproducible in theory but not in practice? For example, if a paper requires tens of thousands of GPUs to reproduce or a large closed-off dataset, then it can only be reproduced in reality by a few large labs.
- For papers which are reproducible both in theory and in practice – how can we ensure that papers published in ICML would actually be able to replicate if such an experiment were attempted?
- What is the best way of publishing the code of the papers so that it is easy for engineers to implement it? Just publishing ipython notebooks it is not sufficient and often hard to make it work in different platforms
- A lot of people tend to understand an algorithm by looking at code and not by following equations. How can we come up with a framework of publishing that includes them. Is pseudocode the best we can do?
- While scientific papers often do an importance analysis of the features, ML papers rarely do proper attribution on the importance of algorithmic components and hyperparameters. What is the best way to “unit-test” an algorithm and do attribution of the results to certain components and hyperparameters
- What does it mean for a paper to have successful or unsuccessful replications?
- Of the papers with attempted replications completed, how many have been published?
- What can be done to ensure that as many papers which are reproducible in theory fall into the last category?
- On the reproducibility issue, what can the Machine Learning community learn from other fields?
- Part of ensuring reproducibility of state-of-the-art is ensuring fair comparisons, proper experimental procedures, and proper evaluation methods and metrics. To this end, what are the proper guidelines for such aspects of machine learning problems? How do they differ among subsets of machine learning?
Our aim in the following workshop is to raise the profile of these questions in the community and to search for their answers. In doing so we aim for papers focusing on the following topics:
- Analysis of the current state of reproducibility in machine learning. Some examples of this include experimental-driven investigations as in [1,2,3]
- Investigations and proposals of proper experimental procedure and evaluation methodologies which ensure reproducible and fair comparisons in the novel literature 
- Tools to help improve reproducibility
- Evidence-driven works investigating the importance of reproducibility in machine learning and science in general
- Connections between the reproducibility situation in Machine Learning and other fields
- Rigorous replications, both failed and successful, of influential papers in the Machine Learning literature.
- With the emergence of new fast prototyping systems such as TensorFlow, CNTK, PyTorch, MXNet, etc it is now much easier to present an implementation, but this is just the beginning. How can we build tools on top of them so that we can get an X-Ray of the algorithm that shows how the components work together.
This workshop likely is relevant and interesting to participants of all co-located conferences: IJCAI-ECAI, AAMAS, and ICML. Reproducibility of research is something that affects most (if not all) scientific fields and is important to emphasize in all the co-located foci and fields. We are targeting ML Researchers/Practitioners from the industry and academia, who want to accelerate transition of research to industrial applications and try to reimplement the ideas from papers as a baseline for their own research.
Call for Papers
We will accept both short paper (4 pages) and long paper (8 pages) submissions. A few papers may be selected as oral presentations, and the other accepted papers will be presented in a poster session. There will be no proceedings for this workshop, however, upon the author’s request, accepted contributions will be made available in the workshop website. Submission are single-blind, peer-reviewed on OpenReview, and open to already published work.
We are also encouraging submissions of reproducible implementations of ML papers in ipython notebook format. The ipython notebook must be self sufficient and it must have very detailed comments and notes along with visuals that demonstrate how equations and pseudo code from the paper translates into code. Authors are encouraged to use small datasets (real and synthetic) to limit the runtime. Further instructions about how to demonstrate the code of the paper can be found here. If the model size is too big and training time is lengthy, it is ok to have pretrained models stored on a website and have the notebook download it. The authors are free to select the python platform of their choice (TensorFlow, PyTorch, CNTK, MxNet, Keras, etc). The submissions will be judged based on the technical clarity and ease of understanding of the code. Amazon will provide cloud containers (Possibly SageMaker) to permanently host the notebooks.
- Submission Deadline: June 5th 2018 23:59 EST
- Acceptance Notifications June 15th 2018
- Camera ready July 1st 2018
 Lucic, Mario, Karol Kurach, Marcin Michalski, Sylvain Gelly, and Olivier Bousquet. “Are GANs Created Equal? A Large-Scale Study.” arXiv preprint arXiv:1711.10337 (2017).
 Melis, Gábor, Chris Dyer, and Phil Blunsom. “On the state of the art of evaluation in neural language models.” arXiv preprint arXiv:1707.05589 (2017).
 Henderson, Peter, Riashat Islam, Philip Bachman, Joelle Pineau, Doina Precup, and David Meger. “Deep reinforcement learning that matters.” arXiv preprint arXiv:1709.06560 (2017).
 Nie, Xinkun, Xiaoying Tian, Jonathan Taylor, and James Zou. “Why adaptively collected data have negative bias and how to correct for it.” (2017).
Nan Rosemary Ke (MILA)
Alex Lamb (MILA)
Peter Henderson (McGill )
Anirudh Goyal (MILA)
Aaron Courville (MILA)
Chris Pal (Politechnique Montreal)
Hugo Larochelle (Google Brain)
Oriol Vinyals (DeepMind)
Yoshua Bengio (MILA)
Nikolaos Vasiloglou (MLTrain)
Alex Dimakis (UT Austin)
George Georgoulas (Lulea Technical University)
Kristian Kersting (TU Darmstadt University)
Hongyang Li: (The Chinese University of Hong Kong)