EN11_Assignments - BC Learning Network

Note: Final exams include literature analysis, but students are also asking for reflective articles on the best works.

I participated in a detailed learning course such as artificial neural network, convolution neural network, supervised learning by recurrent neural network, self organizing map, Boltzmann machine, unsupervised learning with auto encoder etc. In this course, we use the deep learning library Tensorflow and PyTorch to create models in the programming language Python. This is what I have learned in this process. Before I attended this course, I listened to buzzwords and saw headlines - "AI! Machine learning! Robots are coming!" But I think that these mechanisms I do not know what it does. I have discovered that many different learning models are being developed. Each has its own purpose. Adopting a generalized approach, neural networks and deep learning can be modeled as mimicking connections between neurons and neurons in the brain.

The 3rd generation neural network, spiking neural network aims to fill the gap between neuroscience and machine learning by using biologically realistic neuron model for calculation. The spiking neural network (SNN) is fundamentally different from the neural network known to the machine learning community. SNN works with spikes. Spikes are not continuous values, but are individual events occurring at some point. The appearance of spikes is determined by differential equations representing various biological processes. The most important one is the membrane potential of neurons. Basically, when a neuron reaches a certain potential, it will soar and the potential of the neuron will be reset. The most common model is the Leaky integrated ignition (LIF) model. In addition, SNN is usually loosely coupled and uses a special network topology.