Notes

[!PDFs]+ PDFs (2022) Elia's Lecture Notes: 📄 ETH - Deep Learning in Artificial & Biological Neuronal Networks - PDF.pdf (2019) TAs Lecture Notes: 📄 HS2019_LearningInDeepArtificialAndBiologicalNeuronalNetworks.pdf

• Introduction

  • Lecture Notes: ETH/ETH - Deep Learning in Artificial & Biological Neuronal Networks/Lecture Notes - ETH Deep Learning in Artificial & Biological Neuronal Networks/Introduction
  • Extracted Topics:
    • Course Overview
    • Human Brain - Deep Networks Analogies

• Plasticity in the Brain

  • Lecture Notes: Plasticity in the Brain
  • Extracted Topics:
    • Synaptic Plasticity
    • The Hippocampus as a Model System to Study Neural Plasticity
    • Non-Synaptic Plasticity

• Training Methods for Deep ANNs

  • Lecture Notes: Training Methods for Deep ANNs
  • Extracted Topics:
    • The Backpropagation of the Error Method (BP)
    • Feedback Alignment (FA)
    • Target Propagation
    • Local (layer-wise) Training for Deep Neural Networks
    • The Deep Feedback Control Method

• Learning Rules

  • Lecture Notes: Learning Rules
  • Extracted Topics:
    • Why (local) Neuronal Learning Rules are Important?
    • Perceptron Learning Rule
    • ADALINE & Delta Learning Rule
    • Hebbian Learning Rule
    • Oja's Learning Rule
    • Covariance Learning Rule
    • Sanger's Learning Rule
    • Calcium Rule
    • Sejnowski's Infomax Network (ICA) Rule & Bienenstock-Cooper-Monroe (BCM) Rule
    • Triplet Rule (Pfister & Gerstner)
    • Extensions of Hebbian Learning Rules (Neo-Hebbian)

• Reinforcement Learning

  • Lecture Notes: Reinforcement Learning
  • Extracted Topics:
    • Introduction - Reinforcement Learning & the Brain
    • Rescorla - Wagner Rule
    • Temporal Difference Rule & Q-Learning
    • Basic Components of Reinforcement Learning - Policy & Value Functions
    • Markov Chains (MC), Markov Reward Processes (MRPs) & Markov Decision Processes (MDPs)
    • Bellmann (Expectation) Equation
    • Deep Reinforcement (Q) Learning

• Un- and Self-Supervised Learning

  • Lecture Notes: Un- and Self-Supervised Learning
  • Extracted Topics:
    • Unsupervised Learning - Introduction & Motivation
    • Unsupervised Learning (UL) in the Brain
    • Sparse Coding & Relation to Neuroscience
    • Non-Probabilistic UL - PCA, ICA (Infomax)
    • Non-Probabilistic UL - Autoencoders & Supervised Autoencoders
    • Non-Probabilistic UL - Contracting Autoencoders
    • Non-Probabilistic UL - Denoising & Sparse Autoencoders
    • Non-Probabilistic UL - "Homomorphism" Autoencoders
    • Non-Probabilistic UL - Competitive Network Learning
    • Probabilistic (Generative) Unsupervised Learning
    • Probabilistic (Generative) UL - Boltzmann Machines
    • Probabilistic (Generative) UL - Contrastive Divergence
    • Self-Supervised Learning - Pixel-RNN

• Meta-Learning

  • Lecture Notes: Meta-Learning
  • Extracted Topics:
    • Meta-Learning with ANNs - What is Meta-Learning?
    • Meta-Learning with ANNs - Metric-Based (Prototypical, Siamese, Matching and Relation Networks)
    • Meta-Learning with ANNs - Model-Based (Meta & Hyper Networks)
    • Meta-Learning with ANNs - Optimization-Based (Model-Agnostic Meta-Learning)
    • Meta-Learning in the Brain

• Continual Learning

  • Lecture Notes: Continual Learning
  • Extracted Topics:
    • Continual Learning - Introduction
    • Continual Learning - Strategies
    • Continual Learning - Regularization Methods (Elastic Weight Consolidation & Synaptic Intelligence)
    • Continual Learning - Data Replay Methods
    • Continual Learning & the Brain

• Why Spikes?

  • Lecture Notes: Why Spikes?
  • Extracted Topics:
    • What is a Neuronal Spike
    • Digital vs Non-Digital Communication
    • Non-Spiking Biological Systems & Different Types of Action Potentials
    • How To Measure Spiking Activity in a Biological Neuron
    • Temporal Coding Schemes with Spikes
    • Deep Learning with "Time to First Spike"
    • Neuronal Spiking Dynamics (Hodgkin-Huxley Model)

• Deep Learning with Spikes

  • Lecture Notes: Deep Learning With Spikes
  • Extracted Topics:
    • Spiking Neuron Models
    • Supervised Learning in Multi-Layer Spiking Networks - Introduction
    • Supervised Learning in Multi-Layer Spiking Networks - Seq2Seq Learning
    • Neuromorphic Hardware & Spiking Neural Networks

• Learning in Recurrent Neuronal Network

  • Lecture Notes: Learning in Recurrent Neuronal Networks
  • Extracted Topics:
    • RNNs in the Brain - Circuit-Level Recurrence - Anatomical & Functional Evidence
    • RNNs in Machine Learning & Back-Propagation Through Time
    • RNNs in Theoretical Neuroscience - Hopfield Networks, Reservoir Computing & Self-Organizing Recurrent Networks (SORN)
    • Long-Short-Term Memory (LSTM) Networks

• Predictive Coding

  • Lecture Notes: Predictive Coding
  • Extracted Topics:
    • Information Coding
    • Temporal Predictions
    • Predictive Coding - Circuits & Learning
    • Predictive Coding - Problems
    • Bayesian Brain & Free Energy Principle

• Neuromorphic Intelligence

  • Lecture Notes: Neuromorphic Intelligence
  • Extracted Topics:
    • Neuromorphic Intelligence - Introduction & Relation to Neuroscience
    • Neuromorphic Engineering Approach
    • Neuromorphic Synapse Analog Circuits
    • Neuromorphic Processors
    • Neuromorphic Pros & Cons
    • Neuromorphic Applications

• Attention Is All You Need

  • Lecture Notes: Attention Is All You Need
  • Extracted Topics:
    • Attention in Neuroscience - Selective Attention & Visual Saliency
    • Attention Models in Machine Learning - Self-Attention, Transformers and GPTs

• How Can Biological Learning Be So Efficient?

  • Lecture Notes: How Can Biological Learning Be So Efficient?
  • Extracted Topics:
    • Ontogenesis, Kolmogorov Complexity & Self-Reinforcing Networks
    • Representation, Perception & Learning