FML-19： Multi-task Learning for Vertical Federated Machine Learning: A Case Study For Cross-Lingual Short-Text Matching
Published in The 1st International Workshop on Federated Machine Learning for User Privacy and Data Confidentiality (FML-19) in conjunction with the 28th International Joint Conference on Artificial Intelligence (IJCAI-19), 2019
Abstract
With the advert of IOT and the increasing usage of smart devices, the generated data has increased exponentially. The capacity of the machine learning approaches, can be improved by using more high quality labeled data, which are often stored in independent devices. Most of the traditional machine learning approaches require centralizing the multiparty data to train a model, using a central server to store the data, with the goal to obtain superior performance. Federated machine learning can be employed to address aforementioned issues without centralizing the data. In this paper, we propose a vertical federated learning framework using multi-task learning. And, our approach employs partly shared model to protect the privacy. In a practical environment, we conduct extensive experiments on the cross-lingual short-text matching task. Results demonstrate the effectiveness of the proposed method.
Authors : Bo Li, Kele Xu, Haibo Mi, Dawei Feng, Huaimin Wang and Yanbo J. Wang
Paper Download
Not Availiable Yet
Code
Not Availiable Yet