We will use numerous online resources and documents throughout the course. The required and recommended textbooks are listed below. The resources directly relevant to topics covered in the course are listed in the course schedule.

Required:

• Machine Learning in Action by Peter Harrington, 2012.

Optional:

• Python Data Science Handbook: Essential Tools for Working with Data, by Jake VanderPlas, 2023.
• Python Data Science Essentials, by Alberto Boschetti and Luca Massaron, 2018.
• Python for Data Analysis, by Wes McKinney, 2017.
• Data Mining, Practical Machine Learning Tools and Techniques, by Ian H. Witten, Eibe Frank, Mark A. Hall, Christopher J. Pal, 2016.

The structure and grading in the class will be centered around 4 assignments and a final project. The assignments will involve Python implementations of selected machine learning techniques and their applications in various domains. They will also involve the use of various Python libraries to perform preprocessing, data exploration/visualization, and data analysis using different data sets. These assignments must be done individually, unless otherwise specified. You may discuss the assignments with others in the class, but you must develop your own solution to the problems in the assignment. Late assignments will be penalized 10% per day (with weekends counting as one day).

The final project will involve either developing and evaluating an application that uses one or more machine learning algorithms to perform a specific task; or they may involve performing a complete analysis of a complex data set using Python tools. The goal of the project is to integrate several concepts covered during the quarter to achieve a more complex task than those explored in the assignments. Students can propose their own project idea (to be approved) and may complete the project either individually or in groups of up to three members. 

The final grade will be determined (tentatively) based on the following components:

• Assignments = 65%
• Final Project = 35%

The general grading scheme will be based on a curve. At the end of the quarter, some adjustments may be made based on overall class performance as well as signs of individual effort. Plusses and minuses will be given at the high/low ends of each grade range.

(DSC 441 and DSC 430) or CSC 480 are prerequisites for this class.

The following issues and topics will be covered throughout the course. Many of these topics will be revisited several times during the course in a variety of contexts.

• Data Mining and Knowledge Discovery
  • The KDD process and methodology
  • Data preparation for knowledge discovery and machine learning
  • Exploratory data analysis
  • Overview of machine learning tasks
  • Review of Python and overview of Python tools for data analysis and visualization
• Supervised Learning
  • Classification and prediction using K-Nearest-Neighbor
  • Probabilistic models:Naïve Bayes
  • Building decision trees
  • Linear regression models and regularization
  • Gradient descent optimization
  • Support vector machines
  • Evaluating predictive models
  • Hyper-parameter tuning
  • Feature selection
  • Ensemble models: bagging and boosting, including Random Forest, AdaBoost, etc.
• Unsupervised Learning
  • Clustering: K-Means, HSCAN, etc.
  • Hierarchical clustering algorihtms
  • Principal Component Analysis and dimensionality reduction
  • Singular Value Decomposition
  • Matrix factroization
• Possible Applications (covered throughout the course)
  • Recommender systems and personalization
  • Document categorization
  • Concept discovery from text
  • Finding groups using social or behavioral data
  • Building predictive models for target marketing
  • Customer or user segmentation
  • Image segmentation