AI, Neural Networks, and ChatGPT

Artificial intelligence (AI) has recently emerged to be a technology revolution that is able to provide benefits beyond traditional rules-based approaches. AI and neural networks are able to overcome the complexities and optimize the performance of communications networks, computer graphics, multimedia and language processing systems, data science, navigation and voice assistants, and numerous applications. Using 228 informative slides, two interesting projects, and Python code, this course will equip participants with the foundational knowledge on the key building blocks of naturally intelligent learning systems and generative AI, including large-scale language models used in the global phenomenon ChatGPT.

It also contains a short quiz. Learning OutcomesOverview of AI and machine learning, and their capabilitiesStudy the design of ChatGPT, Llama, Claude, Gemini, Grok, and Deep SeekNeural network architecture and implementation, including perceptron and adaline training, backpropagation and attractor networks with memory, recurrent networks, biological and competitive learning, self-organizing map, and interpretable modeling using the Kolmogorov-Arnold networkGenerative AI using supervised, reinforcement, unsupervised trainingSpeech, image, and video AI enhancements using recurrent neural network, convolutional neural network (AlexNet, Unet), Markov and diffusion models, and Adam optimizerChatGPT components, including tokenization, embedding, encoding, decoding, language and post processingAbout the InstructorThe instructor has worked at the Georgia Institute of Technology for many years and has been an active speaker for the IEEE and industry. He has trained hundreds of engineers from various companies around the globe.

One of his books that was published by Cambridge University Press was adopted by an AI company as training material. His current research interest is in optimizing media processing (language, speech, audio, video) and wireless systems using AI. He is a senior member of the IEEE.