18-785: Data, Inference and Applied Machine Learning

Instructor: Prof. Patrick McSharry (patrick@mcsharry.net)

Semester: Fall 2023

Lecture:

Recitation:

Course Overview

This course will provide the methods and skills required for utilizing data and quantitative models to automate predictive analytics and make improved decisions. From descriptive statistics to data analysis to machine learning the course will demonstrate the process of collecting, cleaning, interpreting, transforming, exploring, analyzing and modeling data with the goal of extracting information, communicating insights and supporting decision-making. The advantages and disadvantages of linear, nonlinear, parametric, nonparametric and ensemble methods will be discussed while exploring the challenges of both supervised and unsupervised learning. The importance of quantifying uncertainty, statistical hypothesis testing and communicating confidence in model results will be emphasized. The advantages of using visualization techniques to explore the data and communicate the outcomes will be highlighted throughout. Applications will include visualization, clustering, ranking, pattern recognition, anomaly detection, data mining, classification, regression, forecasting and risk analysis. Participants will obtain hands-on experience during project assignments that utilize publicly available datasets and address practical challenges.

Course Syllabus

The full course syllabus contains additional information and can be found in PDF format HERE. 

Outcomes

After completing this course, students should be able to:

• Design an empirical project in response to a specific objective
• Identify and collect relevant data for undertaking the project
• Acquire data and organize it into a structured format
• Visualize data, identify key characteristics and present a summary
• Describe the advantages and disadvantages of different models
• Decide which models are likely to work best for a given application
• Undertake feature construction and feature selection
• Apply machine learning techniques for estimation and evaluation
• Select an optimal model using statistical approaches
• Produce diagnostic information for investigating model properties
• Understand model weaknesses and where assumptions could fail
• Communicate model output and conclusions to end-users

Textbooks

There will be no required textbooks, though we suggest the following to help you to study (all available online):

• (ESL): Elements of Statistical Learning Trevor Hastie, Robert Tibshirani and Jerome Friedman. 
• (TM): Machine Learning, Tom Mitchell.
• (PRML): Pattern Recognition and Machine Learning, Christopher M. Bishop
• (MJ): Machine Learning Algorithms, The SAS Data Science Blog

Piazza

We will use Piazza for class discussions. Please go to the course Piazza site to join the course forum (note: you must use a cmu.edu email account to join the forum). We strongly encourage students to post on this forum rather than emailing the course staff directly (this will be more efficient for both students and staff). Students should use Piazza to: 

• Ask clarifying questions about the course material.
• Share useful resources with classmates (so long as they do not contain homework solutions).
• Look for students to form study groups.
• Answer questions posted by other students to solidify your own understanding of the material.

Academic Integrity

The course Academic Integrity Policy must be followed when doing assignments and on the message boards at all times. Details on ECE's Academic Integrity Policy can be found in the course syllabus and HERE. 

Grading

The grades for this course will be based on students’ performance on seven homework assignments, a final exam and class participation. Homework assignments will be done individually and turned in via Canvas by the designated due date. Late work will be acceptable until 24 hours past the deadline, but it will lose 10%. The assignments will be graded based on both a writing report and code used to achieve results presented in the report. Class participation will be evaluated based on student’s contribution to discussions both in-class and on the Piazza Discussion Board. When posting or reacting to online discussion threads, students are expected to use their own words and the post should be relevant to the topic under discussion. Make sure to introduce, summarize and explain the article in your own words to enlighten the audience on the point the article is making.

The following is the weight distribution of the grades:

Class participation                                5%

Kahoot Quiz                                         2.5%

Piazza Participation                              2.5%

Homework Assignment 1                      5%

Homework Assignment 2                      10%

Homework Assignment 3                      10%

Homework Assignment 4                      10%

Mid-Semester Exam (Multiple Choice)  10%

Homework Assignment 5                      10%

Homework Assignment 6                      12.5%

Homework Assignment 7                      12.5%

Kaggle Project                                       Bonus 2.5%

Final Exam (Multiple Choice)                10%

Teaching Assistants Contact Info

Schedule (Subject to Change)