CSC 5610 - AI Tools and Paradigms

4 lecture hours 0 lab hours 4 credits
Course Description
This course introduces topics, libraries, and methods used in modern data science and machine learning. Subjects include a survey of tools and techniques used to assess, clean, visualize, analyze, and interpret data. Machine learning concepts, models, and software for classification and regression problems are discussed and practiced extensively.
Prereq: (MTH 1120 or MTH 2340 or MTH 5810 ) and (CSC 1120 or equivalent or instructor consent)
Note: This course is open to qualified undergraduate students. Students may not receive credit for CSC 5610 if they have completed CSC 2611 AI Tools and CSC 2621 Intro to Data Science.
Course Learning Outcomes
Upon successful completion of this course, the student will be able to:
  • Use a Jupyter notebook environment to write, structure, run, document (with Markdown), and debug Python programs  
  • Use Series and Data Frames to load, manipulate, and analyze tabular data sets
  • Assess the quality, quantity, and reliability of a data set for a given data science question or task
  • Leverage appropriate libraries to solve non-trivial data cleaning, preparation, integration, and transformation challenges 
  • Employ a plotting library to visualize data  
  • Use exploratory data analysis (e.g., visualizations and statistical tests) to characterize data sets and between pairs of relationships of variables 
  • Analyze and describe machine learning algorithms including (but not limited to) logistic regression, SVMs, and Random Forests 
  • Explain the geometric and algebraic interpretations of linear and non-linear decision boundaries and their relationship to error metrics (e.g., accuracy)
  • Describe implications of feature representations with respect to linear separability 
  • Apply approaches for engineering and evaluating new features from existing data 
  • Encode non-numerical variables in tabular data as numerical features 
  • Use a library to train (apply) machine learning models for (to) classification and regression problems
  • Choose and justify appropriate experimental designs (e.g., test-train split, cross-fold validation) for evaluating predictions
  • Apply and interpret appropriate metrics for evaluating predictions
  • Communicate computational results and interpretations in language appropriate for a technical audience 
  • Identify ethical implications of potential biases in data sets, algorithms, and applications of AI

Prerequisites by Topic
  • Recent programming experience with a modern, object-oriented language (data structure preferred)
  • Ability to perform basic algebra on vectors

Coordinator
Dr. RJ Nowling


Print-Friendly Page (opens a new window)