Abstract
Since the advent of deep learning, solutions for machine learning problems, ranging from image classification to computer vision to natural language processing, have been revolutionized at an unprecedented rate. The data in these datasets are represented in Euclidean space. Nonetheless, there has been a surge in non-Euclidean data and graph applications in recent years, which adds complexities to current machine learning algorithms due to the independence and complicated relations between graph nodes, motivating a new line of research aiming to generalize existing deep learning methods to graph data. In this literature review, I explore five papers in this field.
This post is also available on my colleague’s blog
Abstract
In recent years, reinforcement learning has witnessed substantial progress in domains such as robotics and Atari game-playing. However, RL algorithms face challenges in terms of sample efficiency and exploration in complex and high-dimensional environments. Model-Based Reinforcement Learning represents a promising approach to address these challenges and enhance the efficiency and performance of RL algorithms.
This post is also available on my colleague’s blog
Abstract
In this article, we provide a comprehensive report on how to get started with the Parallax Eddie robot platform with ROS2, and we discuss the problems we encountered during our study.