Machine learning open source software (MLOSS) is one of the cornerstones of open science and reproducible research. Once a niche area for ML research, MLOSS today has gathered significant momentum, fostered both by scientific community, and more recently by corporate organizations. Along with open access and open data, it enables free reuse and extension of current developments in ML. The past mloss.org workshops at NIPS06, NIPS08, ICML10, NIPS13, and ICML15 successfully brought together researchers and developers from both fields, to exchange experiences and lessons learnt, to encourage interoperability between people and projects, and to demonstrate software to users in the ML community.
Continuing the tradition in 2018, we plan to have a workshop that is a mix of invited speakers, contributed talks and discussion/activity sessions. This year’s headline aims to give an insight of the challenges faced by projects as they seek long-term sustainability, with a particular focus on community building and preservation, and diverse teams. In the talks, we will cover some of the latest technical innovations as done by established and new projects. The main focus, however, will be on insights on project sustainability, diversity, funding and attracting new developers, both from academia and industry. We will discuss various strategies that helps promoting gender diversity in projects (e.g. implementing quotas etc.) and how to promote developer growth within a project.
We aim to make this workshop as diverse as possible within the field. This includes a gender balanced speakers, focussing on programming languages from different scientific communities, and in particular most of our invited speakers represent umbrella projects with a hugely diverse set of applications and users (NumFOCUS, openML, tidyverse).
With a call for participation for software project demos, we aim to provide improved outreach and visibility, especially for smaller OSS projects as typically present in academia. In addition, our workshop will serve as a gathering of OSS developers in academia, for peer to peer exchange of learnt lessons, experiences, and sustainability and diversity tactics.
The workshop will include an interactive session to produce general techniques for driving community engagement and sustainability, such as application templates (Google Summer of Code, etc), “getting started” guides for new developers, and a collection of potential funding sources. We plan to conclude the workshop with a discussion on the headline topic.
|08:25 AM||Opening remarks (Intro)|
|08:30 AM||Gina Helfrich, NumFOCUS (Invited talk)|
|09:00 AM||Christoph Hertzberg, Eigen3 (Invited talk)|
|09:30 AM||Joaquin Vanschoren, OpenML (Invited talk)|
|10:00 AM||Sherpa: Hyperparameter Optimization for Machine Learning Models (Poster spotlight)|
|10:05 AM||How to iNNvestigate neural network’s predictions! (Poster spotlight)|
|10:10 AM||mlpack open-source machine learning library and community (Poster spotlight)|
|10:15 AM||Stochastic optimization library: SGDLibrary (Poster spotlight)|
|10:20 AM||Baseline: Strong, Extensible, Reproducible, Deep Learning Baselines for NLP (Poster spotlight)|
|10:25 AM||skpro: A domain-agnostic modelling framework for probabilistic supervised learning (Poster)|
|Franz J Kiraly|
|10:25 AM||Open Fabric for Deep Learning Models (Poster)|
|10:25 AM||PyLissom: A tool for modeling computational maps of the visual cortex in PyTorch (Poster)|
|10:25 AM||Salad: A Toolbox for Semi-supervised Adaptive Learning Across Domains (Poster)|
|10:25 AM||Gravity: A Mathematical Modeling Language for Optimization and Machine Learning (Poster)|
|10:25 AM||Discussion over morning coffee (Break)|
|10:25 AM||Machine Learning at Microsoft with ML.NET (Poster)|
|10:25 AM||Towards Reproducible and Reusable Deep Learning Systems Research Artifacts (Poster)|
|10:25 AM||Why every GBM speed benchmark is wrong (Poster)|
|10:25 AM||McTorch, a manifold optimization library for deep learning (Poster)|
|10:25 AM||Tensorflex: Tensorflow bindings for the Elixir programming language (Poster)|
|10:25 AM||Open Source Machine Learning Software Development in CERN(High-Energy Physics): lessons and exchange of experience (Poster)|
|10:25 AM||Accelerating Machine Learning Research with MI-Prometheus (Poster)|
|10:25 AM||xpandas - python data containers for structured types and structured machine learning tasks (Poster)|
|11:20 AM||Building, growing and sustaining ML communities (Talk)|
|11:40 AM||PyMC's Big Adventure: Lessons Learned from the Development of Open-source Software for Probabilistic Programming (Talk)|
|12:00 PM||Lunch (on your own) (Break)|
|02:00 PM||James Hensman, GPFlow (Invited talk)|
|02:30 PM||Mara Averick, tidyverse (Invited talk)|
|03:00 PM||Afternoon coffee break (Break)|
|03:30 PM||DeepPavlov: An Open Source Library for Conversational AI (Demo)|
|03:50 PM||MXFusion: A Modular Deep Probabilistic Programming Library (Demo)|
|04:10 PM||Flow: Open Source Reinforcement Learning for Traffic Control (Demo)|
|04:30 PM||Reproducing Machine Learning Research on Binder (Demo)|
|04:50 PM||Panel discussion (Discussion panel)|
|05:30 PM||Closing remarks (Outro)|