How can we make models more robust and more rapidly adaptable, to increase the world's resilience? 

In the face of the dramatic changes provoked by the current pandemic, many machine learning-based systems fail to adapt quickly. The re/insurance industry, with many of its insurance products based on AI models, is particularly impacted. Dynamically changing environments dramatically affect the risk landscape of their portfolios.

This track aims to bring together practitioners from academia, industry (not exclusively re/insurance) and startups, to present and discuss the broad topic of modelling in non-stationary environments, where dynamic modelling is required to cope with an intrinsically changing context that also reacts to the actions taken by the model. Monitoring and addressing model performance degradation in complex enterprise-level machine learning systems will also be a focus of the track, with an emphasis toward AI safety approaches to guarantee models' fairness, interpretability and robustness

We are interested in, and not only, the following sub-topics:

1. Agent-based modelling
2. Hybrid model-based and statistical modelling
3. Reinforcement Learning
4. AI Safety
5. Causal inference

We invite you to submit your contributions for the AMLD 2021: AI & Resilience in dynamic environments track.
Help us to make this track a great interdisciplinary session in one of the largest machine learning and AI events in Europe!

About your presentation

• Language: all presentations are in English
• Duration: we would like to keep the sessions interesting, full of energy and fast-paced, thus we’re looking for talks of a length between 10 and 15 minutes. We will be on a tight schedule and will enforce the time limits rigorously. We suggest that you time your presentation in advance.
• Audience: your public will be made of researchers, practitioners, students, with different levels of knowledge and expertise. We suggest not to dig into very technical details. Still, feel free to provide references and pointers in case someone is interested in investigating further!
• Focus: the best presentations have a point. And one point only. Keep both your talk and its proposal concise and focused. A rule of thumb is the following: if you need more than three paragraphs for your proposal, try to reduce its scope. This will help us to deal with the massive amount of submissions we get, help you to streamline your presentation, and help the audience retain the most out of your talk.
• Originality: one of the things we want to do at AMLD is to push the community forward. Thus, we favor original content. If you want to present a topic that you have talked about elsewhere, try to add a twist, new research or development to it – something to make it unique.

Selection process

Here’s how presentations will be selected:

1. All submissions are anonymized, so that there is no bias towards the submitter.
2. The evaluation takes into account the following criteria:
   1. relevance of the topic to the ML community and for the track
   2. coherence and clarity of the proposal
   3. novelty/originality of the topic
   4. consistency with the time limits
3. The top submissions are then de-anonymized so we can take speaker details into account.

Video recording for presentations

We plan to record and publish all talks online for free, along with a recording of the slide deck, live demo, and any on-presenter-screen activity. We do this for the benefit of the larger ML community and those who couldn’t make it to the conference. Since you retain full ownership of your slides and recording, we’d like to ask you to make your materials and recording available under a Creative Commons (we default to non-commercial reuse) or other open source licenses.

We hope you agree with us, but if you are uncomfortable with this in any way, let us know.

If in doubt, reach out! If you have an idea for a presentation but are unsure, please talk to us! Reach out to info@appliedmldays.org.

Successful applicants will be notified by September 15, 2021 that their proposal has been accepted and they will be given detailed information on the track.

Apply now by filling this form