Bart Kosko

Bart Kosko is affiliated with the University of Southern California in the United States. Their research spans primarily the field of Computer Science, with a strong emphasis on Artificial Intelligence through numerous publications. Additional subfields include Electrical and Electronic Engineering, Cognitive Neuroscience, Computer Vision and Pattern Recognition, and Statistical and Nonlinear Physics.

Their work addresses several main topics in the broader domain of neural network research and related areas. These topics include:

• Neural Networks and Applications
• Advanced Memory and Neural Computing
• Machine Learning and Extreme Learning Machines (ELM)
• Fuzzy Logic and Control Systems
• Cognitive Science and Mapping
• Neural Dynamics and Brain Function
• Advanced Neural Network Applications

Kosko's publication record includes papers in various notable venues. Frequent publication outlets are:

• arXiv (Cornell University)
• IEEE Transactions on Systems Man and Cybernetics Systems
• 2021 20th IEEE International Conference on Machine Learning and Applications (ICMLA)
• Neural Networks
• Neurocomputing

Recent papers authored or coauthored by Kosko highlight their research focus. Selected works include:

• "Noise can speed backpropagation learning and deep bidirectional pretraining" (2020), published in Neural Networks
• "Bidirectional Associative Memories: Unsupervised Hebbian Learning to Bidirectional Backpropagation" (2021), published in IEEE Transactions on Systems Man and Cybernetics Systems

Collaborations play a role in Kosko's research activities. Frequent coauthors include:

• Olaoluwa Adigun
• Akash Kumar Panda
• Kartik Audhkhasi
• Osonde Osoba

The research areas covered by Kosko contribute to the study and advancement of neural networks, memory models, learning algorithms, and cognitive computational systems. Their work encompasses both theoretical and applied aspects within these topics, often exploring mechanisms for improving neural learning methods and architectures.