Machines that learn and teach seamlessly

Abstract

This paper describes an investigation into creating agents that can learn how to perform a task by observing an expert, then seamlessly turn around and teach the same task to a less proficient person. These agents are taught through observation of expert performance and thereafter refined through unsupervised practice of the task, all on a simulated environment. A less proficient human is subsequently taught by the now-trained agent through a third approach-coaching, executed through a haptic device. This approach addresses tasks that involve complex psychomotor skills. A machine-learning algorithm called PIGEON is used to teach the agents. A prototype is built and then tested on a task involving the manipulation of a crane to move large container boxes in a simulated shipyard. Two evaluations were performed-a proficiency test and a learning rate test. These tests were designed to determine whether this approach improves the human learning more than self-experimentation by the human. While the test results do not conclusively show that our approach provides improvement over self-learning, some positive aspects of the results suggest great potential for this approach.

Learning SQL with Artificial Intelligent Aided Approach

“…Although in early Computer Aided Instruction (CAI) systems students may have had some influence on navigation through the curriculum, they all received the same contents [3], [6], [7]. In the later CAI systems branching provided different responses to a student's answer, depending on what student's response was [6], [8].…”

“…In the later CAI systems branching provided different responses to a student's answer, depending on what student's response was [6], [8]. This kind of CAI systems possessed no domain knowledge, meaning that every feedback had to be provided by experts manually.…”

“…The most acclaimed ITS systems evolved through the history are cognitive tutors [11], [12] followed by approaches, based on constraint-based modelling [3], [13], [14] or the principle of constructing student models with machine learning techniques [6], [15].…”

“…Building ITS systems has always been expensive and time-consuming process, where many different experts with adequate programming knowledge have been involved [5], [6], [13], [23]. We would like to find a balance among the complexity of building such systems, invested time and their price.…”

Imitation Learning for Object Manipulation Based on Position/Force Information Using Bilateral Control

“…Humans can easily adapt to perturbations in working conditions but considerable research into designing robots that can learn human manipulations is still required. Technique have been actively conducted [5] [6] [7]. Recently, some studies have significantly reduced the number of trials by imitating human manipulative skills via remote control [8].…”