DEEP LEARNING

---

• An artificial neural network (ANN) is the piece of a computing system designed to simulate the way the human brain analyzes and processes information. It is the foundation of AI and solves problems that would prove impossible or difficult by human or statistical standards. ANNs have self-learning capabilities that enable them to produce better results as more data becomes available.
• Video Link: Neural Networks and Deep Learning- Andrew Ng 
• Project Link: Github
• Important Functions: a. Gradient Descent

                                                     b. Sigmoid, ReLU, LeakyReLU 

                                                     c. Forward & Backward Propagation 

                                                     d. Xavier Gorat, He init

                                                     e. Normal & Uniform Distribution

                                                     g. Sochastic Gradient Descent

                                                     h. Adagrad Optimizer, Adadelta & RMSProp

                                                     i. ADAM Optimizer

                                                     j. Softmax Regression

• Computer vision is a field of Artificial Intelligence that trains computers to interpret and understand the visual world. Using digital images from cameras and videos and DeepLearning models, machines can accurately identify and classify objects and then react to what they “see.” 
•  In DeepLearning, a convolutional neural network (CNN, or ConvNet) is a class of Deep Neural Networks, most commonly applied to analyzing visual imagery. They are also known as shift invariant or space invariant artificial neural networks SIANN, based on their shared-weights architecture and translation invariance characteristics. They have applications in image & video processing, recommender system, image classification, medical image analysis, & many more.

                                     a. Digital Image Processing

                                     b. Digital Video Processing

• Video Link:  Fundamentals of Computer Vision - Mubarak Shah 

                                      Convolutional Neural Networks - Andrew Ng

• Important Functions: For Computer Vision

                                                                a. Filtering & OpticalFLow

                                                                b. Canny Edge Detector

                                                                c. Harris Detector & SIFT Transform

                                                                d. Laplace & Gaussian Theorem

                                                      For Convolution Neural Network

                                                                a. Padding & Pooling

                                                                b. Transfer Learning

                                                                c. Data Augmentation

                                                                d. Localization , Landmark detection & Object detection

                                                                e. Nonmax Suppression & Anchor Boxes

• CNN Features: a. Neural Style Transfer

                                          b. YOLO Algorithm

                                          c. R-CNN

                                          d. Mask R-CNN  

• Project Link:  TrafficSignClassifier for SelfDrivingCar 

                                        Vehicle Detection & Tracking SelfDrivingCar 

                                        LaneDetection for SelfDrivingCar 

                                        Mask-RCNN for SelfDrivingCar          

• An artificial neural network ANN is the piece of a computing system designed to simulate the way the human brain analyzes and processes information. It is the foundation of AI and solves problems that would prove impossible or difficult by human or statistical standards. ANNs have self-learning capabilities that enable them to produce better results as more data becomes available.
• Video Link:  Recurrent Neural Networks | Sequence Models  

                                     NLP with DeepLearning

• Important Functions: a. RNN

                                                     b. BackPropogation Equation

                                                     c. Language Modelling

                                                     d. Vanishing gradients with RNN

                                                     e. GRU

                                                     f. Long Short Term Memory

                                                     g. Text Summarization

• Project Link: Image Caption Generator

                                       seq2seq English to French Translation

• Electroencephalography EEG is an electrophysiological monitoring method to record the electrical activity of the brain. It is typically noninvasive, with the electrodes placed along the scalp, although invasive electrodes are sometimes used, as in electrophotography, sometimes called intracranial EEG.
•   A generative adversarial network (GAN) is a class of Machine Learning frameworks designed by Ian Goodfellow and his colleagues in 2014. Two neural network contest with each other in a game (in the form of a zero-sum game, where one agent's gain is another agent's loss). 
• Video Link:  Generative Adversarial Networks (GANs) - Computerphile

                                      Generative Adversarial Networks (GANs) - Ahlad Kumar

• Project Link: FaceGenerating GAN

                                       Pokemon GAN

• Research Oriented Project: 

                                      a.  Epileptic Seizure Detection Based on EEG Signals

                                       b. Visual Reconstruction of Image from Spoken Word using EEG

                                                         Abstract Research Paper pdf

•  Reinforcement learning (RL) is an area of Machine Learning concerned with how software agent ought to take actions in an environment in order to maximize the notion of cumulative reward. Reinforcement learning is one of three basic machine learning paradigms, alongside supervised learning and unsupervised learning.
• Video Link: Introduction to Reinforcement Learning- DeepMind
• Important Functions: a. State & Information State

                                                     b. Markov Decision Process

                                                     c. Policies & Value Function 

                                                     d. Q-Learning

                                                     e. Bellman Optimality Equation

                                                     f. Q value and Updation 

• Project Link: DeepQ, DuelingDeepQRL on gym LunarLander