Explainable Fuzzy AI Challenge 2022: Winner’s Approach to a Computationally Efficient and Explainable Solution

Abstract

An explainable artificial intelligence (XAI) agent is an autonomous agent that uses a fundamental XAI model at its core to perceive its environment and suggests actions to be performed. One of the significant challenges for these XAI agents is performing their operation efficiently, which is governed by the underlying inference and optimization system. Along similar lines, an Explainable Fuzzy AI Challenge (XFC 2022) competition was launched, whose principal objective was to develop a fully autonomous and optimized XAI algorithm that could play the Python arcade game “Asteroid Smasher”. This research first investigates inference models to implement an efficient (XAI) agent using rule-based fuzzy systems. We also discuss the proposed approach (which won the competition) to attain efficiency in the XAI algorithm. We have explored the potential of the widely used Mamdani- and TSK-based fuzzy inference systems and investigated which model might have a more optimized implementation. Even though the TSK-based model outperforms Mamdani in several applications, no empirical evidence suggests this will also be applicable in implementing an XAI agent. The experimentations are then performed to find a better-performing inference system in a fast-paced environment. The thorough analysis recommends more robust and efficient TSK-based XAI agents than Mamdani-based fuzzy inference systems.

2D-Localization Based on Tracking Antenna Using Artificial Intelligence Mapping (AIM) Approach

“…Thus, a distinction between the proposed modified triangulation method and the traditional triangulation method can be observed. In the MTM, each position group consists of three nodes (positions) and is determined by three known tracking antenna positions within the range of coupling between the object and tracking antenna beam pattern: (A 1 , A 2 , A 3 ), (A 4 , A 5 , A 6 ), … (A n-2 , A n-1 , A n ), as well as the measured angles between them, as determined from the object [38]. The angle measurements for the three nodes with respect to the reference orientation  were ( 1 ,  2 ,  3 ), ( 4 ,  5 ,  6 ), … ( n- The most crucial simplification involves converting the world coordinate frame into one of the nodes, solving the issue related to that node, and adding node coordinates to the calculated object position.…”

“…Starting with a set of membership functions for each input and a set for each output, a fuzzy logic system was designed. The membership functions are then subjected to a set of restrictions to produce a "crisp" output value [38]. It should be noted that membership functions do not necessarily have triangular shapes.…”

Deepfakes Generation and Detection: A Short Survey

“…Nonetheless, present deepfake or face manipulation detection techniques cannot distinguish the intent. For deepfake detection framework’s interpretability and dependability, various advanced combinations of techniques such as fuzzy inference systems [ 187 ], layer-wise relevance propagation [ 188 ], and the Neural Additive Model [ 189 ] could be helpful.…”

Mathematical Fuzzy Logic in the Emerging Fields of Engineering, Finance, and Computer Sciences

“…Mishra et al [12] explored the area of explainable artificial intelligence (XAI) based on the background of the Explainable Fuzzy AI Challenge 2022. The authors proposed a fully autonomous and optimized XAI algorithm that can play the Python arcade game "Asteroid Smasher".…”