Introduction To Deep Learning

---

October 7 through October 28, 2023. Saturdays 1:30-4:30pm.  

The IEEE North Jersey Section Communications Society (ComSoc chapter) is offering a course entitled "INTRODUCTION TO DEEP LEARNING". 

Deep learning is a transformative field within artificial intelligence and machine learning that has revolutionized our ability to solve complex problems in various domains, including computer vision, natural language processing, and reinforcement learning. This introductory course on deep learning is designed to provide students with an understanding how these amazing successes are made possible by drawing inspiration from the way that brains, both human and otherwise, operate. Students will gain a comprehensive foundation in the principles, techniques, and applications of deep neural networks.

The IEEE North Jersey Section's Communications Society Chapter can arrange for providing IEEE CEUs - Continuing Education Units (for a $5 charge) upon completion of the course.  Course prices: $75 for Undergrad/Grad/Life/ComSoc members, $100 for IEEE members, $150 for non-IEEE members

---

---

---

---

Topics and agenda:
1. Fundamental Concepts: Explore the fundamental concepts of artificial neural networks, backpropagation, activation functions, and gradient descent, laying the groundwork for deep learning understanding.

Familiarize students with popular deep learning frameworks such as TensorFlow and PyTorch, enabling hands-on experience in model development and training.

Examine a variety of neural network architectures, including feedforward networks, convolutional neural networks (CNNs) and recurrent neural networks (RNNs).

2. Training Deep Neural Networks: Study techniques for training deep neural networks effectively, including optimization algorithms, weight initialization, regularization, and dropout.

Learn data preprocessing techniques and best practices for preparing datasets for deep learning tasks, including data augmentation and normalization.

3. Computer Vision Applications: Apply deep learning to computer vision problems, including image classification, object detection, and image generation using generative adversarial networks (GANs).

4. Natural Language Processing (NLP): Explore how deep learning is used in NLP tasks such as sentiment analysis, machine translation, and text generation.

This course assumes a basic understanding of machine learning concepts and programming skills in Python. Familiarity with linear algebra and calculus will be beneficial, but not mandatory.  Statistical software (Python, Scikit-learn) and Deep Learning Frameworks (Pytorch, TensorFlow) will be used throughout the course for the exploration of different learning algorithms and for the creation of appropriate graphics for analysis.

Learning objectives:  Subjects covered include these and other deep learning related materials: artificial neural networks, training deep neural networks, RNN, CNN, image recognition, natural language processing, GANs, data processing techniques, and NN architectures.

The course is intended to be subdivided into 3-hour sessions. Each lecture is further subdivided into lecture, guided and independent project based exercises to build experience with hands-on techniques.  This course will be held at FDU - Teaneck, NJ campus.  Checks should NOT be mailed to this address.  Can bring checks in person or use online payments at registration.  Email the organizer for any questions about course, registration, or other issues.

Technical Requirements: Students will need access to the Python programming language. In addition to a standard Python installation, most programming exercises will use the package Scikit-learn.  Basic programming skills and some familiarity with the Python language are assummed.
Students are expected to be able to bring a laptop onto which most of these libraries can be pre-installed using python's pip install.

Most of the coding in this course will use the Python programming language. Coding examples and labs will be distributed in the form of Juypter notebooks. In addition to standard Python, most programming exercises will use either the PyTorch or TensorFlow libraries.

---