Skip to content

Latest commit

 

History

History
70 lines (47 loc) · 3.35 KB

18763.md

File metadata and controls

70 lines (47 loc) · 3.35 KB
Raw

18-763/14-763: Systems and Tool Chains for AI Engineers

Category Difficulty (out of 5)
Homework Assignments 3
Exam 3
Course Project 3
In-class quizzes 2

The premise of this course is to build a broad and solid foundation in Artificial Intelligence Infrastructure that will pay significant dividends throughout a student’s research and work career across data science and Artificial Intelligence related fields. In class the following topics are covered:

  • Data Collection and Storage.
  • Data Streaming.
  • Data Engineering.
  • Modern ML Frameworks.
  • Model Validation and Monitoring.
  • Neural Network Design and Implementation.
  • Embedded Machine Learning.
  • Deployment of ML Models to the Cloud

Logistics

  • Two lectures per week.
  • Only one exam - Final Exam.
  • Grading will be based on assignments, final exam, in-class quizzes, course project.
  • This class is traditionally offered in Fall.

Pre-requisites

Preferred qualifications include practical experience in Python, with a strong preference for prior exposure to machine learning. Students lacking Python proficiency will have the opportunity to attend a Python refresher lecture during the second week of the class.

Class structure

Lectures will be synchronous, and attendance is required. Lecture slides are delivered via TopHat, an online course delivery system. TopHat can be accessed on a smartphone or laptop. An in-class quiz is conducted after every lecture, and the code for that is provided at the end of the class (which is why attendance is required).

Learning Resources

There is no primary textbook, as most reading material will come from research papers and other technical documentation.

Topics Covered

Some of the main topics covered in the course are:

  • SQL, NoSQL, SparkSQL
  • AWS DynamoDB
  • Confluent Kafka
  • SparkML
  • Pytorch and Tensorflow
  • TinyML

Assesments

Students are encouraged to attend class regularly, read the assigned reading material and participate in class discussions. The final grade will be based upon 1 exam, 1 project, 8 homework assignments, and in-class quizzes. In-class participation will grant student extra-credit that will help boosting their quiz score.

Final Exam: is an open-note test

Quizzes: are offered during each lecture via Canvas. Each quiz will be accessible via a unique access code that will be provided to students of the corresponding section. Students will have 5 minutes to answer 1-2 multiple choice questions

Assignments: will provide the opportunity to practice the concepts that are taught during the class

Project: details are released in week 3. Each student will have the option to team up with another student for the project and you will choose one of two project options to submit

How to do well

  • Having hands-on experience in Python is highly preferred, and some exposure to machine learning is a plus. For those needing a Python refresher, there will be a lecture in the second week of the class.
  • Although the assignmnets are not difficult, there will be something due almost every week, make sure to do them on time as there are no late days for the course. There is a 20% penalty for late submission.
  • Have a lot of fun! You will learn a lot. I really enjoyed this course.