Lecture 09. CNNs - clairedavid/ml_in_hep GitHub Wiki

The lecture format

Learn how to learn with video.

  • Guest speaker.

Outline (extract keywords and put them in intro)

  1. Definition of CNNs: A brief explanation of what CNNs are and how they differ from traditional neural networks.
  2. Motivation for using CNNs: A discussion of why CNNs are particularly well-suited for tasks involving images, such as image classification, object detection, and segmentation.
  3. Overview of CNN architecture: A high-level description of the main components of a CNN, including the convolutional layers, pooling layers, and fully connected layers.
  4. Convolutional layers: A detailed explanation of how convolutional layers work, including the concept of convolutional filters and stride.
  5. Pooling layers: A description of the purpose and function of pooling layers in a CNN, including the two main types of pooling (max pooling and average pooling).
  6. Fully connected layers: A discussion of the role of fully connected layers in a CNN, and how they are used to make final predictions based on the output of the convolutional and pooling layers.
  7. Applications of CNNs: Examples of real-world tasks that CNNs have been successfully applied to, such as image classification, object detection, and segmentation.

List of Videos

  1. "Convolutional Neural Networks (CNNs) - The Math of Intelligence (Week 4)" by Siraj Raval: This video provides a concise and intuitive explanation of CNNs, including the concept of convolutional filters and stride, and how CNNs are used for image classification.
  2. "Convolutional Neural Networks (CNNs) - Deep Learning with Neural Networks and TensorFlow" by Sentdex: This video covers the basics of CNNs, as well as more advanced topics such as transfer learning and fine-tuning.
  3. "Convolutional Neural Networks (CNNs) Explained" by Data School: This video provides a thorough overview of CNNs, including the different types of layers and how they work together to process images and make predictions.
  4. "Convolutional Neural Networks (CNNs) for Computer Vision" by Computer Science: This video covers the basics of CNNs, as well as more advanced topics such as object detection and segmentation.

These are just a few examples, and there are many other good videos available online that explain CNNs in different ways. It may be helpful to watch a few different videos to get a well-rounded understanding of this topic.