International Conference on Computer and Knowledge Engineering

Home / 15th International Conference on Computer and Knowledge Engineering

Adaptive Prioritization in Experience Replay Using Feedback from Multiple Learning Signals

Authors :

Seyed Hossein Mostafavi¹ Mohammad Bagher Naghibi Sistani²

1- Ferdowsi university of mashhad 2- Ferdowsi university of mashhad

Keywords :

Deep reinforcement learning،experience replay،prioritized experience replay،multi-criteria sampling،adaptive weighting

Abstract :

Deep reinforcement learning (DRL) has made significant progress in recent years. Many DRL algorithms utilize experience replay to store past experiences and reuse them during training. One main challenge in this process is choosing which experiences to sample. While most previous methods rely on one or two sampling criteria, this study introduces a method that incorporates four distinct criteria, with their weights adaptively tuned based on environmental feedback. Simulation results demonstrate that the proposed method outperforms previous approaches in various reinforcement learning environments.

List of archived papers

Multi-Layer Collaborative Graph with BPR Similarity Embedding for Recommender System

Mostafa Ghorbani - Azadeh Mansouri

IR-LPR: Large Scale of Iranian License Plate Recognition Dataset

Mahdi Rahmani - Melika Sabaghian - Seyyedeh Mahila Moghadami - Mohammad Mohsen Talaie - Mahdi Naghibi - Mohammad Ali Keyvanrad

SingAll: Scalable Control Flow Checking for Multi-Process Embedded Systems

Mehdi Amininasab - Ahmad Patooghy - Mahdi Fazeli

Robat-e-Beheshti: A Persian Wake Word Detection Dataset for Robotic Purposes

Parisa Ahmadzadeh Raji - Yasser Shekofteh

Multi-source Ensemble Model for Scene Recognition

Amir Hossein Saleknia - Ahmad Ayatollahi

Automated Person Identification from Hand Images\\using Hierarchical Vision Transformer Network

Zahra Ebrahimian - Seyed Ali Mirsharji - Ramin Toosi - Mohammad Ali Akhaee

Design and Simulation of a Low PDP Full Adder by Combining Majority Function and TGDI Technique in CNTFET Technology

Mahsa Mohammadi

An Interactive Approach for Query-based Multi-Document Scientific Text Summarization

Mohammadsadra Nejati - Azadeh Mohebi - Abbas Ahmadi

Attention Transfer in Self-Regulated Networks for Recognizing Human Actions from Still Images

Masoumeh Chapariniya - Sara Vesali Barazande - Seyed Sajad Ashrafi - Shahriar B.Shokouhi

Impossible differential and zero-correlatin linear cryptanalysis of Marx, Marx2, Chaskey andSpeck32

Mahshid Saberi - Nasour Bagheri - Sadegh Sadeghi

more

Samin Hamayesh - Version 43.7.0