Auto AI Seminars (WP7)

We are pleased to announce the outline of the sixth event in our TAILOR WP7 seminar series.

Please let us know if you plan to participate by entering your name in this Seminar 6 main doc [TAILOR WP7].

The seminar will consist of two parts: a scientific and a social part. Below you can find a detailed overview of each part for the fourth seminar on 27 September from 15:30-17:00.


Scientific part: Zoom meeting: https://universiteitleiden.zoom.us/j/61008083679?pwd=TjNXcHN5RnllYjRLS3ovY1ZxMUZaUT09 – Meeting ID: 610 0808 3679 – Passcode: i!&gA1Xn

15:30-16:00 — Talk by Dragi Kocev—Meta learning from a comprehensive empirical study of multi-label classification methods

Abstract:

We present our recent work on understanding and explaining the performance of multi-label classificaion (MLC) methods across a wide range of MLC datasets. Essentially, we perform a comprehensive meta-learning study of MLC methods and datasets. Several studies provide surveys of methods and datasets for MLC, and a few provide empirical comparisons of MLC methods. However, they are limited in the number of methods and datasets considered. Here, we analyze 40 MLC data sets by using 50 meta features describing different properties of the data. The main findings of this study are as follows. First, the results of the analysis identify RFPCT, RFDTBR, ECCJ48, EBRJ48, and AdaBoost.MH as the best-performing methods across the spectrum of performance measures. Second, the most prominent meta features that describe the space of MLC data sets are the ones assessing different aspects of the label space. Third, the meta models show that the most important meta features describe the label space, and, the meta features describing the relationships among the labels tend to occur a bit more often than the meta features describing the distributions between and within the individual labels. Fourth, the optimization of the hyperparameters can improve the predictive performance, however, quite often the extent of the improvements does not always justify the resource utilization.

16:00-16:30 — Round table updates for all WP7 tasks

the expectation is that each TAILOR partner with funding allocated under WP7 sends at least one representative who can give a brief update. It is understood that there may not be a substantive update on each task at every meeting. Each update should focus on highlighting the following:

  • Recent and ongoing work, potential collaborations
  • Collaboration between different tasks;
  • Collaboration with other work packages;
  • New collaborations (with people with whom that group has never worked with before).

Social part: GatherTown 16:30-17:00

Submit a PDF by email to a.w.bosman@liacs.leidenuniv.nl if what you want to present something)

  • Suggestions for improving the seminar format
  • Informal chats about possible collaborations
  • Brainstorm/discuss ideas and mechanism for making our WP7 more impactful, more efficient and/or more enjoyable
  • Chat about anything else you feel worthwhile!

Hope to see you at the seminar!

Adem, Holger, Koen and Manuel (the TAILOR WP7 Seminar Organising Committee)

Date

27 Sep 2022
Expired!

Time

15:30 - 17:00

Location

Online

Organizer

TAILOR WP7 Seminar Organising Committee
Email
a.w.bosman@liacs.leidenuniv.nl