Youtor
Lecture 10 Training Neural Networks I
>> YOUR LINK HERE: ___ http://youtube.com/watch?v=lGbQlr1Ts7w
Lecture 10 discusses many of the nuts-and-bolts details you need to think about when designing and training neural networks. We discuss different activation functions, and see how the classical sigmoid and hyperbolic tangent activations lead to vanishing gradients while more recent alternatives like rectified linear units (ReLU) mitigate this problem. We discuss data preprocessing and weight initialization and see how Xavier and MSRA initialization schemes make deep networks easier to train. We discuss common strategies for regularizing neural networks including dropout. We see how many regularizers fit into a common framework of adding noise during training, then averaging out the noise at test-time. Through this lens we discuss data augmentation, DropConnect, fractional pooling, stochastic depth, cutout, and Mixup. • Slides: http://myumi.ch/0WeP9 • _________________________________________________________________________________________________ • Computer Vision has become ubiquitous in our society, with applications in search, image understanding, apps, mapping, medicine, drones, and self-driving cars. Core to many of these applications are visual recognition tasks such as image classification and object detection. Recent developments in neural network approaches have greatly advanced the performance of these state-of-the-art visual recognition systems. This course is a deep dive into details of neural-network based deep learning methods for computer vision. During this course, students will learn to implement, train and debug their own neural networks and gain a detailed understanding of cutting-edge research in computer vision. We will cover learning algorithms, neural network architectures, and practical engineering tricks for training and fine-tuning networks for visual recognition tasks. • Course Website: http://myumi.ch/Bo9Ng • Instructor: Justin Johnson http://myumi.ch/QA8Pg
#############################
