MLDublin Goes Remote

For our third remote we have two speaker who are presenting upcoming papers.
I've like to think that MLDublin is a friendly audience for people to present to. We're excited that we have a two presentations that are for upcoming conferences.

AGENDA:

[18:40 - 19:00] Getting Online

[19:00 - 19:10] Welcome

[19:10 - 19:30] Prcheta Sen, PhD Student @ ADAPT Centre, DCU

Towards Socially Responsible AI: Cognitive Bias-Aware Multi-Objective Learning

Abstract: Human society had a long history of suffering from cognitive biases leading to social prejudices and mass injustice. The prevalent existence of cognitive biases in large volumes of historical data can pose a threat of being manifested as unethical and seemingly inhumane predictions as outputs of AI systems trained on such data. To alleviate this problem, we propose a bias-aware multi-objective learning framework that given a set of identity attributes (e.g. gender, ethnicity etc.)and a subset of sensitive categories of the possible classes of prediction outputs, learns to reduce the frequency of predicting certain combinations of them, e.g. predicting stereotypes such as ‘most blacks use abusive language’, or ‘fear is a virtue of women’. Our experiments conducted on an emotion prediction task with balanced class priors shows that a set of baseline bias-agnostic models exhibit cognitive biases with respect to gender, such as women are prone to be afraid whereas men are more prone to be angry. In contrast, our proposed bias-aware multi-objective learning methodology is shown to reduce such biases in the predictid emotions.

[19:30 - 19:50] Tong Jian, PhD Student @ Department of Electrical and Computer Engineering, Northeastern University

Learn-Prune-Share for Lifelong Learning

Abstract: In lifelong learning, we wish to maintain and update a model (e.g., a neural network classifier) in the presence of new classification tasks that arrive sequentially. In this paper, we propose a Learn-Prune-Share (LPS) algorithm which addresses the challenges of catastrophic forgetting, parsimony, and knowledge reuse simultaneously. LPS splits the network into task-specific partitions via an ADMM-based pruning strategy. This leads to no forgetting, while maintaining parsimony. Moreover, LPS integrates a novel selective knowledge sharing scheme into this ADMM optimization framework. This enables adaptive knowledge sharing in an end-to-end fashion. Comprehensive experimental results on two lifelong learning benchmark datasets are provided to demonstrate the effectiveness of our approach. Our experiments show that LPS consistently outperforms multiple state-of-the-art competitors.

Tong Jian is currently pursuing a Ph.D. degree in the Department of Electrical and Computer Engineering, Northeastern University. She works under the guidance of Prof. Stratis Ioannidis in the field of machine learning. Her current research efforts are focused on investigating the application of Machine Learning for radios and Lifelong Learning.

This event is strictly for Machine Learning professionals, researchers and students only

• If you finally can't make it, please RSVP to "NO" as soon as possible so that other people can take your place.

• Please, make sure that you agree with our Attendance Policy (https://www.meetup.com/Machine-Learning-Dublin/about/)

• If you are interested in speaking or sponsoring this event, please contact any of the organisers.