Learn · Machine Learning for Engineering and science applications
Lecture 48 of 107Week 5
Summary of Week 05
Prof. Balaji Srinivasan Prof. Ganapathy Krishnamurthi · IIT Madras
Credits and licence
- From
- Machine Learning for Engineering and science applications, IIT Madras (at NPTEL)
- Licence
- CC BY-NC-SA, as NPTEL states it. Non-commercial use only.
- Attribution
- NPTEL states: "Distributed under Creative Commons Attribution-ShareAlike / CC BY - NC - SA", beside a BY-NC-SA badge.
- Who can watch
- Tweens profiles and older. How courses are placed
All the lectures
- Introduction to the Course History of Artificial Intelligence
- Overview of Machine Learning
- Why Linear Algebra ? Scalars, Vectors, Tensors
- Basic Operations
- Norms
- Linear Combinations Span Linear Independence
- Matrix Operations Special Matrices Matrix Decompositions
- Introduction to Probability Theory Discrete and Continuous Random Variables
- Conditional, Joint, Marginal Probabilities Sum Rule and Product Rule Bayes' Theorem
- Bayes' Theorem - Simple Examples
- Independence Conditional Independence Chain Rule Of Probability
- Expectation
- Variance Covariance
- Some Relations for Expectation and Covariance (Slightly Advanced)
- Machine Representation of Numbers, Overflow, Underflow, Condition Number
- Derivatives,Gradient,Hessian,Jacobian,Taylor Series
- Matrix Calculus (Slightly Advanced)
- Optimization - 1 Unconstrained Optimization
- Introduction to Constrained Optimization
- Introduction to Numerical Optimization Gradient Descent - 1
- Gradient Descent - 2 Proof of Steepest Descent Numerical Gradient Calculation Stopping Criteria
- Introduction to Packages
- The Learning Paradigm
- A Linear Regression Example
- Linear Regression Least Squares Gradient Descent
- Coding Linear Regression
- Generalized Function for Linear Regression
- Goodness of Fit
- Bias-Variance Trade Off
- Gradient Descent Algorithms
- Introduction to Week 5 (Deep Learning)
- Logistic Regression
- Binary Entropy cost function
- OR Gate Via Classification
- NOR, AND, NAND Gates
- XOR Gate
- Differentiating the sigmoid
- Gradient of logistic regression
- Code for Logistic Regression
- Multinomial Classification- Introduction
- Multinomial Classification - One Hot Vector
- Multinomial Classification - Softmax
- Schematic of multinomial logistic regression
- Biological neuron
- Structure of an Artificial Neuron
- Feedforward Neural Network
- Introduction to back prop
- Summary of Week 05
- Introduction to Convolution Neural Networks (CNN)
- Types of convolution
- CNN Architecture Part 1 (LeNet and Alex Net)
- CNN Architecture Part 2 (VGG Net)
- CNN Architecture Part 3 (GoogleNet)
- CNN Architecture Part 4 (ResNet)
- CNN Architecture Part 5 (DenseNet)
- Train Network for Image Classification
- Semantic Segmentation
- Hyperparameter optimization
- Transfer Learning
- Segmentation of Brain Tumors from MRI using Deep Learning
- Activation Functions
- Learning Rate decay, Weight initialization
- Data Normalization
- Batch Norm
- Introduction to RNNs
- Example - Sequence Classification
- Training RNNs - Loss and BPTT
- Vanishing Gradients and TBPTT
- RNN Architectures
- LSTM
- Why LSTM Works
- Deep RNNs and Bi- RNNs
- Summary of RNNs
- Introduction.
- Knn
- Binary decision trees
- Binary regression trees
- Bagging
- Random Forest
- Boosting
- Gradient boosting
- Unsupervised learning & Kmeans
- Agglomerative clustering
- Probability Distributions- Gaussian, Bernoulli
- Covariance Matrix of Gaussian Distribution
- Central Limit Theorem
- Naïve Bayes
- MLE Intro
- PCA-part 1
- PCA-part 2
- Support Vector Machines
- MLE, MAP and Bayesian Regression
- Introduction to Generative model
- Generative Adversarial Networks (GAN)
- Variational Auto-encoders (VAE)
- Applications: Cardiac MRI - Segmentation & Diagnosis
- Applications: Cardiac MRI Analysis - Tensorflow code walkthrough
- Introduction to Week 12
- Application 1 description - Fin Heat Transfer
- Application 1 solution
- Application 2 description - Computational Fluid Dynamics
- Application 2 solution
- Application 3 description - Topology Optimization
- Application 3 solution
- Application 4 - Solution of PDE/ODE using Neural Networks
- Summary and road ahead
- Live Session 30-08-2019