Deep Learning with TensorFlow Training in San Francisco

Syllabus of Deep Learning with TensorFlow Training in San Francisco

• 1. Introduction to Deep Learning
• 2. Introduction to Numpy
• 3. Introduction to Tensorflow and Keras

• 1. Solution of Equations, row and column Interpretation
• 2. Vector Space Properties
• 3. Partial Derivative of Polynomial and Two conditions for Local Minima
• 4. Physical Interpretation of gradient (Direction of Maximum Change)
• 5. Matrix Vector Multiplication
• 6. EVD and interpretation of Eighen Vectors
• 7. Linear Independence and Rank of Matrix
• 8. Orthonormal Matrices, Projection Matrices, Vandemonde Matrix, Markov Matrix, Symmetric, Block Diagonal

• 1. Intuition behind Linear Regression, classification
• 2. Grid Search
• 3. Gradient Descent
• 4. Training Pipeline
• 5. Metrics ROC Curve, Precision Recall Curve
• 6. Calculating Entropy

• 1. Evolution of Perceptrons, Hebbs Principle, Cat Experiment
• 2. Single layer NN
• 3. Tensorflow Code
• 4. Multilayer NN
• 5. Back propagation, Dynamic Programming
• 6. Mathematical Take on NN
• 7. Function Approximator
• 8. Link with Linear Regression
• 9. Dropout and Activation
• 10. Optimizers and Loss Functions

• 1. 1D and 2D Convolution
• 2. Why CNN for Images and speech?
• 3. Convolution Layer
• 4. Coding Convolution Layer
• 5. Learning Sharpening using single convolution Layer in Tensor-Flow

• 1. Convolution
• 2. Pooling
• 3. Activation
• 4. Dropout
• 5. Batch Normalization
• 6.Object Classification
• 7. Creating Batch in Tensorflow and Normalize
• 8. Training MNIST and CIFAR datasets
• 9. Understanding a pre-trained Inception Architecture
• 10. Input Augmentation Techniques for Images

• 1. Finetuning last layers of CNN Model
• 2. Selecting appropriate Loss
• 3. Adding a new class in the last Layer
• 4. Making a model Fully Convolutional for Deployment
• 5. Finetune Imagenet for Cats vs Dog Classification.

• 1. Different types of problem in Objects
• 2. Difficulties in Object Detection and Localization
• 3. Fast RCNN
• 4. Faster RCNN
• 5. YOLO v1-v3
• 6. SSD
• 7. MobileNet

• 1. Image Compression Simple Autoencoder
• 2. Denoising Autoencoder
• 3. Variational Autoencoder and Reparematrization Trick
• 4. Robust Word Embedding using Variational Autoencoder

• 1. Evolution of Recurrent Structures
• 2. LSTM, RNN, GRU, Bi-RNN, Time-Dense
• 3. Learning a Sine Wave using RNN in Tensorflow
• 4. Creating Autocomplete for Harry Potter in Tensorflow

• 1. Generative vs Discrimative Models
• 2. Theory of GAN
• 3. Simple Distribution Generator in Tensorflow using MCMC (Markov Chain Monte Carlo)
• 4. DCGAN,WGANs for Images
• 5. InfoGANs, CycleGANs and Progressive GANs
• 6. Creating a GAN for generating Manga Art

• 1. Model Free Prediction
• 2. Monte Carlo Prediction and TD Learning
• 3. Model Free Control with REINFORCE and SARSA Learning
• 4. Assignment : Implementation of REINFORCE and SARSA Learning in Gridworld
• 5. Off policy vs On Policy Learning
• 6. Importance Sampling for Off Policy Learning
• 7. Q Learning

• 1. Understanding Deep Learning as Function Approximator
• 2. Theory of Behavioral Cloning and Deep Q Learning
• 3. Revisiting Point Collector Example in Unity and
• 4. Assignment : Training Cartpole Example via Deep Q Learning

• 1. Face Detection using Yolo-v3
• 2. Building Autocomplete Feature using RNNs
• 3. Real-time Depth Prediction and Pose Estimation
• 4. How is Deep Learning used in Autonomous Driver Assistant systems
• 5. Tips and Tricks for scaling and easy Deployment of Deep Learning Models