Inline training: a technique for continuous, within-task learning
نویسندگان
چکیده
منابع مشابه
Multi-task Learning for Continuous Control
Reliable and effective multi-task learning is a prerequisite for the development of robotic agents that can quickly learn to accomplish related, everyday tasks. However, in the reinforcement learning domain, multi-task learning has not exhibited the same level of success as in other domains, such as computer vision. In addition, most reinforcement learning research on multitask learning has bee...
متن کاملA B-Learning Model for Training within Electrical Tests Domain
An electrical test performing involves high risk therefore utility companies require high qualified electricians. Traditionally, training on electrical tests has been based on classroom courses; and recently it has been supported by virtual reality systems. These systems have improved training and reduced training time. However the training still depends on courses schedule and instructors, and...
متن کاملMulti-Task Deep Reinforcement Learning for Continuous Action Control
In this paper, we propose a deep reinforcement learning algorithm to learn multiple tasks concurrently. A new network architecture is proposed in the algorithm which reduces the number of parameters needed by more than 75% per task compared to typical single-task deep reinforcement learning algorithms. The proposed algorithm and network fuse images with sensor data and were tested with up to 12...
متن کاملThe Maze Task: Training Methods for Second Language Learning
The maze task was created for psycholinguistic experimental testing (Forster et al., 2009). However, this paper explores the merits of this task as a language training program for beginning Spanish learners. The attributes of providing ample comprehensible input and immediate corrective feedback allow the maze task to be considered as a potential supplemental pedagogical tool. Moreover, transfe...
متن کاملMulti-Task Learning for Improved Discriminative Training in SMT
Multi-task learning has been shown to be effective in various applications, including discriminative SMT. We present an experimental evaluation of the question whether multi-task learning depends on a “natural” division of data into tasks that balance shared and individual knowledge, or whether its inherent regularization makes multi-task learning a broadly applicable remedy against overfitting...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Research in Learning Technology
سال: 2018
ISSN: 2156-7077
DOI: 10.25304/rlt.v26.1994