STEM Program

Artificial Intelligence: Machine Learning with Python

Faculty Advisor: Professor, Mathematics, Georgia Institute of Technology

Research Practicum Introduction

While the ideas behind AI date back to the mid-1950s, its popularity and practical use has become pervasive only in the last decade or so. To be powerful, AI needs large amounts of data, lots of computational power and memory, and efficient algorithms. With advances in technology, computer science and mathematics, these three requirements were finally met in the last decade. AI is finding its place in most fields and everyday activities. This explosion is only accelerating, as is reflected in the popularity of ChatGPT and the impressive growth of the value of Nvidia. Some would argue that learning the basics of AI is necessary in most fields and gives a competitive advantage in any field. This is what this program experience offers. 

Machine Learning, the main aspect of AI, is the subject of this research experience. If you participate, you will learn both the fundamentals and pragmatic aspects of machine learning. The fundamentals mean the ideas, the calculations (mathematics) the software does behind the scenes, and the reasons those calculations are done. This fundamental understanding is fundamental to be a good AI/Machine Learning/Data scientist that will have an edge over the competition.

You will also have the opportunity to learn some of the programing language Python and how to use the software used in Machine Learning. Some models will be created in the meetings, as examples, but you will learn to create your own Machine Learning model. To that end, you will select a dataset (project) within two weeks of the first meeting. Once you are done with your project, you will write a paper explaining your findings.

 Possible Topics Covered in Meetings:

  • Introduction to Data Science and Machine Learning: Understanding supervised learning, data organization, and model errors.

  • Model Parameters and Optimization: Minimizing model errors, avoiding underfitting/overfitting, and introduction to neural networks.

  • Programming Essentials for Data Science: Introduction to Python, key libraries (Numpy, Pandas, Keras), and data handling.

  • Building a Machine Learning Model: Steps from data pre-processing to training and validating a model in Python.

  • Real-World Machine Learning Applications: Implementing continuum, binary classification, and multiclassification problems in code. 

Possible Topics For Final Project: 

  • Machine Learning in Healthcare: Create a model that predicts the onset of diabetes based on diagnostic measures such as weight, height, blood pressure, etc

  • Predicting House Price: Create a model that prices houses according to different factors like: number of rooms, area, crime rates, etc

  • Machine Learning in Admission Policies: Data on students of a college, including data before they were admitted and data on how they did during their college years (such as GPA, years to graduation, etc), can be used as input of an ML model to develop a model that can later be used to predict how new applicants will do - a powerful tool for Admissions Officers to decide on admission.   

  • Object Recognition: Create a model that recognizes the objects in images

  • Machine Learning and Fake News: Create a model that detects if a news article is fake.

  • Machine Learning in Cancer Diagnosis: Create a model that can diagnose patients with breast cancer or not, from information obtained from images from cell nuclei from the tumors.

  • Or other topics in this subject area that you are interested in, and that your professor approves after discussing it with you.

Program Detail

  • Cohort Size: 3-5 students

  • Duration: 12 weeks

  • Workload: Around 4 hours per week (including class and homework time)

  • Target Students: 9-12th grade students interested in Mathematics, Artificial Intelligence, Machine Learning, and/or Interdisciplinary STEM studies. 

  • Prerequisites: No prior knowledge is required. The students will be provided with videos and other material (that includes notes and code) to supplement their learning.

  • Schedule: TBD (meetings will take place for around one hour per week, with a weekly meeting day and time to be determined a few weeks prior to the class start date)