我要吃瓜

I'm a PhD student currently working on an investigation into explaining the decisions of commonly-used non-deterministic solvers for optimization problems, this is a joint project between the 我要吃瓜, Robert Gordon University and BT. The project is centred around Explainable AI, while this is well established as a concept the current research success has primarily been focused on methods that mimic human reasoning, meaning the path to solution can be more readily understood by end users and decision makers. In the case of non-deterministic solvers, the journey to a solution is driven by much more inherently random processes that can gleam and 'store' problem learning as they solve as opposed to making deductions from prior experience or knowledge. I will focus on surrogate problem models to investigate new ways of generating user-understandable problem knowledge from analysis of algorithm behaviour, as well as look at and investigate the use of natural language generation and visualisation to convert any insight gained into comprehensible explanations for domain experts and decision makers