Auflistung nach Schlagwort "Transfer learning"
1 - 8 von 8
Treffer pro Seite
Sortieroptionen
- ZeitschriftenartikelAutomated Transfer for Reinforcement Learning Tasks(KI - Künstliche Intelligenz: Vol. 28, No. 1, 2014) Bou Ammar, Haitham; Chen, Siqi; Tuyls, Karl; Weiss, GerhardReinforcement learning applications are hampered by the tabula rasa approach taken by existing techniques. Transfer for reinforcement learning tackles this problem by enabling the reuse of previously learned behaviours. To be fully autonomous a transfer agent has to: (1) automatically choose a relevant source task(s) for a given target, (2) learn about the relation between the tasks, and (3) effectively and efficiently transfer between tasks. Currently, most transfer frameworks require substantial human intervention in at least one of the previous three steps. This discussion paper aims at: (1) positioning various knowledge re-use algorithms as forms of transfer, and (2) arguing the validity and possibility of autonomous transfer by detailing potential solutions to the above three steps.
- ZeitschriftenartikelCognitive Complexity and Analogies in Transfer Learning(KI - Künstliche Intelligenz: Vol. 28, No. 1, 2014) Ragni, Marco; Strube, GerhardThe ability to learn often requires transferring relational knowledge from one domain to another. It is difficult for humans and computers to identify the respective source domain from which relational characteristics can be applied to the target domain. An additional source of human reasoning difficulty is the complexity of the transformation function. In this article we investigate two domains in which the identification of relational patterns and of a transformation function are necessary: number series and geometrical analogy problems. Characteristics of the human processes are presented and existing cognitive models are discussed.
- ZeitschriftenartikelEfficient Supervision for Robot Learning Via Imitation, Simulation, and Adaptation(KI - Künstliche Intelligenz: Vol. 33, No. 4, 2019) Wulfmeier, MarkusRecent successes in machine learning have led to a shift in the design of autonomous systems, improving performance on existing tasks and rendering new applications possible. Data-focused approaches gain relevance across diverse, intricate applications when developing data collection and curation pipelines becomes more effective than manual behaviour design. The following work aims at increasing the efficiency of this pipeline in two principal ways: by utilising more powerful sources of informative data and by extracting additional information from existing data. In particular, we target three orthogonal fronts: imitation learning, domain adaptation, and transfer from simulation.
- ZeitschriftenartikelMulti-phase Fine-Tuning: A New Fine-Tuning Approach for Sign Language Recognition(KI - Künstliche Intelligenz: Vol. 36, No. 1, 2022) Sarhan, Noha; Lauri, Mikko; Frintrop, SimoneIn this paper, we propose multi-phase fine-tuning for tuning deep networks from typical object recognition to sign language recognition (SLR). It extends the successful idea of transfer learning by fine-tuning the network’s weights over several phases. Starting from the top of the network, layers are trained in phases by successively unfreezing layers for training. We apply this novel training approach to SLR, since in this application, training data is scarce and differs considerably from the datasets which are usually used for pre-training. Our experiments show that multi-phase fine-tuning can reach significantly better accuracy in fewer training epochs compared to previous fine-tuning techniques
- ZeitschriftenartikelRegularization-Based Multitask Learning With Applications to Genome Biology and Biological Imaging(KI - Künstliche Intelligenz: Vol. 28, No. 1, 2014) Widmer, Christian; Kloft, Marius; Lou, Xinghua; Rätsch, GunnarThe aim of multitask learning is to improve the generalization performance of a set of related tasks by exploiting complementary information about the tasks. In this paper, we review established approaches for regularization based multitask learning, sketch some recent developments, and demonstrate their applications in Computational Biology and Biological Imaging.
- ZeitschriftenartikelRobots Learn Increasingly Complex Tasks with Intrinsic Motivation and Automatic Curriculum Learning(KI - Künstliche Intelligenz: Vol. 35, No. 1, 2021) Nguyen, Sao Mai; Duminy, Nicolas; Manoury, Alexandre; Duhaut, Dominique; Buche, CedricMulti-task learning by robots poses the challenge of the domain knowledge: complexity of tasks, complexity of the actions required, relationship between tasks for transfer learning. We demonstrate that this domain knowledge can be learned to address the challenges in life-long learning. Specifically, the hierarchy between tasks of various complexities is key to infer a curriculum from simple to composite tasks. We propose a framework for robots to learn sequences of actions of unbounded complexity in order to achieve multiple control tasks of various complexity. Our hierarchical reinforcement learning framework, named SGIM-SAHT, offers a new direction of research, and tries to unify partial implementations on robot arms and mobile robots. We outline our contributions to enable robots to map multiple control tasks to sequences of actions: representations of task dependencies, an intrinsically motivated exploration to learn task hierarchies, and active imitation learning. While learning the hierarchy of tasks, it infers its curriculum by deciding which tasks to explore first, how to transfer knowledge, and when, how and whom to imitate.
- ZeitschriftenartikelTowards Learning of Generic Skills for Robotic Manipulation(KI - Künstliche Intelligenz: Vol. 28, No. 1, 2014) Metzen, Jan Hendrik; Fabisch, Alexander; Senger, Lisa; Gea Fernández, José; Kirchner, Elsa AndreaLearning versatile, reusable skills is one of the key prerequisites for autonomous robots. Imitation and reinforcement learning are among the most prominent approaches for learning basic robotic skills. However, the learned skills are often very specific and cannot be reused in different but related tasks. In the project 'Behaviors for Mobile Manipulation', we develop hierarchical and transfer learning methods which allow a robot to learn a repertoire of versatile skills that can be reused in different situations. The development of new methods is closely integrated with the analysis of complex human behavior.
- ZeitschriftenartikelTransfer for Automated Negotiation(KI - Künstliche Intelligenz: Vol. 28, No. 1, 2014) Chen, Siqi; Ammar, Haitham Bou; Tuyls, Karl; Weiss, GerhardLearning in automated negotiation is a difficult problem because the target function is hidden and the available experience for learning is rather limited. Transfer learning is a branch of machine learning research concerned with the reuse of previously acquired knowledge in new learning tasks, for example, in order to reduce the amount of learning experience required to attain a certain level of performance. This paper proposes a novel strategy based on a variation of TrAdaBoost—a classic instance transfer technique—that can be used in a multi-issue negotiation setting. The experimental results show that the proposed method is effective in a variety of application domains against the state-of-the-art negotiating agents.