Nature-Inspired AI Algorithms
As the plethora of applications for Artificial Intelligence keeps growing, so does the collection of tools and algorithms used to address their respective problems. Based on observations in nature, a class of novel algorithmic approaches has emerged in recent years, proving to be extremely powerful. For example, deep neural networks revolutionized machine learning by modeling how the brain passes information between neurons. Likewise, genetic algorithms inspired by Darwin's theory of natural evolution have solved problems in fields like engineering, social sciences, and medicine. Additionally, the designs of robots specialized for navigating specific terrains have been based on the anatomy of real-life organisms, including dogs, spiders, snakes, octopuses, and humans.
Prof. Bari is a lead researcher in yet another promising field of nature-inspired AI, called Swarm Intelligence. This branch of research studies the logic underlying swarm behavior such as the coordinated movements of flocking birds or schools of fish, the self-organized intelligence of ant colonies or bee hives, or the flight of bats using echolocation. The simulation of this behavioral logic in the computer can then serve as the basis for new algorithms capable of addressing complex problems in domains such as finance, healthcare, marketing, and education.
Selected Publications:
Abdelghani Bellaachia, Anasse Bari. "Flock by leader: a novel machine learning biologically inspired clustering algorithm." In International Conference in Swarm Intelligence, pp. 117-126. Springer, Berlin, Heidelberg, 2012.
Abdelghani Bellaachia, Anasse Bari. "A flocking based data mining algorithm for detecting outliers in cancer gene expression microarray data." In 2012 International Conference on Information Retrieval & Knowledge Management, pp. 305-311. IEEE, 2012.
Abdelghani Bellaachia, Anasse Bari. "SFLOSCAN: A biologically-inspired data mining framework for community identification in dynamic social networks." In 2011 IEEE Symposium on Swarm Intelligence, pp. 1-8. IEEE, 2011.
