computer, ai, artificialintelligence, bigdata, datascience, techtraining, techschool, onlineeducation, technology, aiaaplication, cstu, masterdegree, computerscience data, jobseeker, pytorch, deeplearning, nlp, neuralnetworks, programming
CSE590/MB590 Special Topics (1.5 credits)
  - Deep Learning with PyTorch

This course is an introduction to deep learning with a focus on its application in computer vision. Deep learning is a branch of machine learning which mainly uses the technology of neural networks.
  • » 23 hours (8 weeks) in class lecturing plus dedicated mentoring sessions from our faculty of industry experts
  • » 1.5 semester credits for both certificate and master’s degree
  • » Access to high-quality live class recording
  • » Online live classroom available for all classes
  • » Lifetime learning resources for our students
  • $ 990
Course Description

This course is an introduction to deep learning with a focus on its application in computer vision. Deep learning is a branch of machine learning which mainly uses the technology of neural networks. We will discuss the basics of computer vision, machine learning and venture into the deep learning theories and applications. We will also learn a variety of machine learning and deep learning frameworks with a focus on PyTorch. The introduction to basic neural networks, convolutional neural networks and recurrent neural networks is combined with the development of actual applications in the computer vision space.

Prerequisite: Working experience, basic math knowledge, programming experience with Python.

Course Objectives
  • Demonstrate a basic understanding of computer vision, machine learning and deep learning concepts.
  • Explain the key components of computer vision algorithms and practices.
  • Describe the major concepts of machine learning, its main framework derived from traditional data analytics
  • Describe the connection between machine learning and deep learning
  • Explain the development of machine learning and deep learning frameworks.
  • Demonstrate ability to build model, train model, tune model using deep learning technology on specific computer vision projects.
Course Textbook
  • Deep Learning with Python, by François Chollet, published by Manning Publications, ISBN 9781617294433
  • Deep Learning with PyTorch, by E. Stevens and L. Antiga, published by Manning Publications, ISBN 9781617295263

CSTU does not have a student bookstore. Students are required to purchase textbooks required for their courses on the open market. In accordance with the current HEOA requirements, CSTU will provide the ISBN and retail price of our texts along with information on various purchasing options and buyback programs. The ISBN and price information are provided in the syllabus. Course materials can be purchased from any source, the CSTU website offers a convenient means of obtaining required course materials. CSTU cautions students about obtaining course materials from overseas sources because of the risk of delivery time and quality of the materials. Purchase decisions should not be based on the purchase price alone

Course Topics

Week 1-3: Machine Learning

  • Basic concepts and math
    • Introduction to classification problems
    • Perceptron Algorithm
    • Error function: sigmoid vs. softmax
    • Maximum Likelihood
    • Cross Entropy
  • Introduction to Jupyter Notebook
  • Logistic Regression
  • Linear Regression
  • Gradient Descent
  • Neural Network Basics
    • Architecture
    • Feedforward
  • Backpropagation
  • Introduction to ML/DL frameworks with a focus on PyTorch
  • Project 1: Classifying movie reviews

Week 4: Image Processing / Computer Vision

  • Basic Theory and Practices
  • Introduction to openCV
  • Project 2: Image feature extraction

Week 5-8: Deep Learning

  • CNN
    • Concept and applications
    • Strategy and solutions for all challenges within the workflow
    • Project 3: Handwritten digit recognition on MINIST dataset
  • RNN
    • LTSM
    • ResNet
    • Introduction to YOLO v3
    • Project 4: Object detection/recognition - tied with real world application
  • Generative Deep Learning
    • Introduction to Generative Adversarial Networks
Your Instructor

Bo Shen

Bo Shen, Ph.D., Cofounder & CTO of Hayden AI Technologies; a passionate technologist and effective engineering leader with over 20 years of experience from big corporate research labs to all stages of startups. Bo is specialized in developing state-of-the-art technologies, building big scale solutions from the ground up, and driving/managing all aspects of engineering/technology processes. Before cofounding and serving as CTO for Hayden AI, Bo rose from a software architect to the CTO of a dominant OTT video service company called Vuclip. Before that, Bo was a senior researcher with Hewlett-Packard Laboratories, worked in the areas of video encoding/transcoding/streaming, content distribution, and big scale multimedia systems, and was published extensively in many top technical journals and conferences.