期刊
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
卷 15, 期 2, 页码 -出版社
ASSOC COMPUTING MACHINERY
DOI: 10.1145/3434748
关键词
Multi-task learning; deep neural networks; neural architecture search
资金
- Spanish Ministry of Economy, Industry and Competitiveness [TIN2016-78365-R]
- Spanish Ministry of Science and Innovation [PID2019-104966GB-I00]
- Basque Government [IT-1244-19]
- project 3KIA - SPRI-Basque Government through Elkartek [KK-2020/00049]
- University of the Basque Country [PIF16/238]
Multi-task learning involves using a single model to perform multiple similar tasks, expanding performance range. This study introduces heterogeneous tasks in a single learning procedure and develops an illustrative model with classification, regression, and data sampling tasks. The model exhibits good performance, with potential for future research directions.
Multi-task learning, as it is understood nowadays, consists of using one single model to carry out several similar tasks. From classifying hand-written characters of different alphabets to figuring out how to play several Atari games using reinforcement learning, multi-task models have been able to widen their performance range across different tasks, although these tasks are usually of a similar nature. In this work, we attempt to expand this range even further, by including heterogeneous tasks in a single learning procedure. To do so, we firstly formally define a multi-network model, identifying the necessary components and characteristics to allow different adaptations of said model depending on the tasks it is required to fulfill. Secondly, employing the formal definition as a starting point, we develop an illustrative model example consisting of three different tasks (classification, regression, and data sampling). The performance of this illustrative model is then analyzed, showing its capabilities. Motivated by the results of the analysis, we enumerate a set of open challenges and future research lines over which the full potential of the proposed model definition can be exploited.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据