Actual Instructor: Phileas Dazeley-Gaist, Wolfram Research

This workshop will highlight computational applications of Wolfram technologies in geography and environmental science, from accessing remote sensing imagery and building professional GIS projects and visualizations to spatial modeling of environmental phenomena such as forest fires, earthquakes, and volcanic eruptions. (No prior experience in Wolfram Language is required)

•    Customizing your environment
•    Parallel computing on the SCC
•    Jobs monitoring and profiling: CPU and GPU utilization, memory usage
•    Profiling programs for performance optimization
•    General optimization strategies

This is an introduction to the essential features of Python. This first part of the tutorial includes an introduction to basic types, if-statements, functions, lists, dictionaries, loops, and modules. The tutorial includes the use of a popular Python development environment and covers setting up Python on your own computer in addition to using Python on the SCC. This is a two-part tutorial so please remember to sign up for both sessions.

This tutorial is a continuation of "Introduction to Python, Part One" and introduces more features of the language, common libraries such as numpy and matplotlib, and the basics of debugging Python programs. Please make sure you sign up for part one as well.

If you do not have Python installed on your home machine, please read and follow these instructions prior to attending the tutorial.

This is the first part of a two-part tutorial series. Be sure to also register for Part Two to continue building your knowledge.

What to Expect: This session introduces Scikit-Learn, a powerful Python library for machine learning. Scikit-Learn supports supervised and unsupervised learning and offers tools for:

• Data preprocessing
• Model fitting
• Model selection
• Evaluation
• And much more

Through hands-on exercises with real datasets, you'll learn to develop models using modern algorithms, including:

• Linear regression
• Decision trees and random forests
• K-means clustering
• Dimensionality reduction

We'll also provide an overview of the general machine learning workflow and wrap up with guidance on further ML resources.

Preparation: If Python is not installed on your machine, follow these instructions.

A conda environment file with all necessary packages will be shared before the session, along with activation instructions.

Prerequisites: Experience with Python programming using Jupyter Notebook and familiarity with libraries like NumPy, Pandas, and Matplotlib.

Get ready to explore the capabilities of Scikit-Learn and start building practical machine learning solutions!

This is the second part of a two-part tutorial series. Be sure to also register for Part One.

What to Expect: This session introduces Scikit-Learn, a powerful Python library for machine learning. Scikit-Learn supports supervised and unsupervised learning and offers tools for:

• Data preprocessing
• Model fitting
• Model selection
• Evaluation
• And much more

Through hands-on exercises with real datasets, you'll learn to develop models using modern algorithms, including:

• Linear regression
• Decision trees and random forests
• K-means clustering
• Dimensionality reduction

We'll also provide an overview of the general machine learning workflow and wrap up with guidance on further ML resources.

Preparation: If Python is not installed on your machine, follow these instructions.

A conda environment file with all necessary packages will be shared before the session, along with activation instructions.

Prerequisites: Experience with Python programming using Jupyter Notebook and familiarity with libraries like NumPy, Pandas, and Matplotlib.

Get ready to explore the capabilities of Scikit-Learn and start building practical machine learning solutions!

This is the first part of a two-part tutorial on PyTorch. Be sure to also register for Part Two to continue building your knowledge.

What to Expect: This session introduces PyTorch, a popular and versatile Python library for deep learning, optimized for acceleration processing using GPUs. You’ll gain hands-on experience building and training neural networks for binary classification.

• Why PyTorch?
• GPU acceleration using PyTorch Tensors
• PyTorch Autograd for automatic differentiation
• Working with Data
• Datasets and Data Loaders in PyTorch
• Building Neural Networks
• Developing deep learning models for binary classification using PyTorch

• Experience with Python programming, especially using Jupyter Notebook, is required.
• Before the tutorial, ensure Python is installed on your machine. Detailed setup instructions and a conda environment file with the required packages will be shared.
• If you plan to use your own computer, the conda environment must be installed and activated in advance.

• Familiarity with Python NumPy library.
• Basic understanding of machine learning and deep learning concepts and experience in using them.

Additional Recommendation: For those new to machine learning, consider attending the preceding tutorials on Machine Learning with Python Scikit-Learn to build foundational knowledge.

Get ready to dive into PyTorch and create powerful deep learning models!

This is the second part of a two-part tutorial on PyTorch. Be sure to also register for Part One.

What to Expect: This session introduces PyTorch, a popular and versatile Python library for deep learning, optimized for acceleration processing using GPUs. You’ll gain hands-on experience building and training neural networks for binary classification.

• Why PyTorch?
• GPU acceleration using PyTorch Tensors
• PyTorch Autograd for automatic differentiation
• Working with Data
• Datasets and Data Loaders in PyTorch
• Building Neural Networks
• Developing deep learning models for binary classification using PyTorch

• Experience with Python programming, especially using Jupyter Notebook, is required.
• Before the tutorial, ensure Python is installed on your machine. Detailed setup instructions and a conda environment file with the required packages will be shared.
• If you plan to use your own computer, the conda environment must be installed and activated in advance.

• Familiarity with Python NumPy library.
• Basic understanding of machine learning and deep learning concepts and experience in using them.

Additional Recommendation: For those new to machine learning, consider attending the preceding tutorials on Machine Learning with Python Scikit-Learn to build foundational knowledge.

Get ready to dive into PyTorch and create powerful deep learning models!

What to Expect

This tutorial introduces Hugging Face, the leading open-source platform where the machine learning community collaborates on models, datasets, and applications. Best known for its Transformers library for natural language processing, Hugging Face also provides a vibrant ecosystem for sharing, exploring, and deploying machine learning models.

You will gain hands-on experience navigating the Hugging Face platform, exploring models that meet specific criteria, and building tools using these models. By the end, you’ll understand how Hugging Face supports modern AI development and how to leverage its tools effectively in your own projects.

• Introduction to Hugging Face and its role in open-source AI
• Navigating the Hugging Face ecosystem and model hub
• Finding and evaluating models based on specific requirements
• Understanding the platform’s tools, libraries, and community features
• Hands-on practice: selecting models, working with them in Python, and building simple applications

• Prior experience with Python programming (especially Jupyter Notebooks) is required
• Ensure Python is installed on your machine; detailed setup instructions and a conda environment file will be provided in advance
• If you plan to use your own computer, install and activate the conda environment ahead of time

• Familiarity with the NumPy library
• Basic understanding of machine learning and deep learning concepts with some practical experience

If you are new to machine learning, consider attending the preceding tutorials on Machine Learning with Python (Scikit-Learn) and Deep Learning with PyTorch to build a solid foundation.

Get ready to explore Hugging Face, experiment with open-source models, and build your own AI-powered applications!

Python is now widely used for numerical calculations and data analysis. This tutorial is an introduction primarily to the Numpy library which provides data structures and algorithms that are optimized for numeric data. The Numpy library is the basis for a wide variety of numeric and graphics libraries in Python. The usage of the numpy multi-dimensional array type will be covered in detail. The Scipy library and how it can be effectively used with Numpy arrays and other Python data structures will be discussed. This tutorial assumes familiarity with Python.

Prerequisite: If you are new to the Python programming language we strongly recommend that you also register for the "Introduction to Python" two-part tutorial.

If you do not have Python installed on your home machine, please read and follow these instructions prior to attending the tutorial.

This tutorial will introduce the basics of Data Analysis with Python and its powerful libraries such as Pandas and Matplotlib.

•    Importing and Exporting the data
•    Basic data processing, cleaning, and manipulation
•    Basic inferential statistical analysis
•    Data Visualization techniques

In this tutorial we will go over the basic usage of the package “sf”, which provides tools for working with vector spatial data (points, lines, and polygons). We will go over how to read/write GIS vector files, how to work with the attribute table, and introduce some basic GIS spatial functions.

Prerequisites: We expect those who attend this tutorial to be familiar with the basics of programming using R and a basic understand of GIS theory.

If you do not have R and RStudio installed on your laptop, please read and follow these instructions prior to attending the tutorial.

In this tutorial we will go over the basic usage of the package “terra”, which provides tools for working with raster data. We will go over the basic functions of reading, writing, and manipulating raster data.

Prerequisites: We expect those who attend this tutorial to be familiar with the basics of programming using R and a basic understand of GIS theory.

If you do not have R and RStudio installed on your laptop, please read and follow these instructions prior to attending the tutorial.

This tutorial is a hands-on walkthrough of the basics of the popular neuroimaging software package FreeSurfer. This tutorial will cover how to utilize FreeSurfer’s MRI viewing tool freeview, running FreeSurfer’s cortical reconstruction tool "recon-all”, and inspecting the outputs of the recon-all cortical reconstruction using BU’s Shared Computing Cluster (SCC).

GeoPandas is a Python library that extends the functionality of the Pandas DataFrame object. This allows one to import vector data (points, polylines, and polygons) into a DataFrame object but the python library enables additional functions and spatial attributes used for spatial operations on the DataFrame and enables one to generate maps showing the spatial data. In this tutorial I will introduce you to the GeoPandas functions that will allow you to import GIS data, apply select spatial operations, and produce some simple maps.

Prerequisite: Basic programming in Python and using Pandas DataFrame.

This tutorial will cover the basics of running a machine learning analysis — commonly called “multivariate pattern analysis” or MVPA for short — on magnetic resonance imaging (MRI) data using the MATLAB software package CoSMoMVPA. The tutorial will go over some of the theory behind MVPA before diving into the hands-on tutorial where we will replicate the results of a highly cited paper on multivariate representations of faces and objects in the ventral temporal cortex (Haxby et al. 2001).

Note that this is Part One of a two-part tutorial and you should also register for Part Two.

What to Expect

In this hands-on tutorial, participants will gain practical experience building LLM-driven chatbots and Retrieval-Augmented Generation (RAG) systems using Python. The session will introduce core concepts such as large language models, with a clear focus on why and when RAG is needed for real-world applications. Through guided exercises, attendees will implement end-to-end RAG pipelines that enable chatbots to answer questions directly from documents. The emphasis is on applied system design, and experimentation rather than training models from scratch.

Key Topics Covered

• Introduction to LLMs
• Introduction to RAG (Retrieval-Augmented Generation)
• Why RAG is needed?
• How retrieval works: embeddings -> vector databases -> similarity search
• Hands-on practice: Build RAG chatbots that answers questions from documents

• Intermediate to advanced Python programming proficiency is required. You should be comfortable working with Python packages, virtual environments, and reading and modifying existing Python scripts.
• Prior exposure to NLP concepts (tokens, embeddings, transformers) is helpful.
• A basic understanding of machine learning and deep learning concepts with some practical experience.

Recommended Background: If you are new to machine learning, consider attending the preceding tutorials on Machine Learning with Python (Scikit-Learn), Deep Learning with PyTorch, and Open Source Generative AI with Hugging Face to build a solid foundation.

• Researchers, engineers, and data scientists looking to apply LLMs to real document-based tasks.
• Practitioners interested in building RAG-based chatbots for knowledge bases, documentation, or reports.
• Participants with diverse backgrounds (computer science, data science, applied research, domain sciences) who want a practical entry point into Generative AI.

• Those seeking a deep dive into training LLMs from scratch or into the mathematical foundation of NLP or LLMs or optimization theory
• Participants with no prior Python programming experience.

Get ready to explore LLMs, experiment with open-source models, and build your own AI-powered ChatBot!

Note that this is Part Two of a two-part tutorial and you should also register for Part One.

What to Expect

In this hands-on tutorial, participants will gain practical experience building LLM-driven chatbots and Retrieval-Augmented Generation (RAG) systems using Python. The session will introduce core concepts such as large language models, with a clear focus on why and when RAG is needed for real-world applications. Through guided exercises, attendees will implement end-to-end RAG pipelines that enable chatbots to answer questions directly from documents. The emphasis is on applied system design, and experimentation rather than training models from scratch.

Key Topics Covered

• Introduction to LLMs
• Introduction to RAG (Retrieval-Augmented Generation)
• Why RAG is needed?
• How retrieval works: embeddings -> vector databases -> similarity search
• Hands-on practice: Build RAG chatbots that answers questions from documents

• Intermediate to advanced Python programming proficiency is required. You should be comfortable working with Python packages, virtual environments, and reading and modifying existing Python scripts.
• Prior exposure to NLP concepts (tokens, embeddings, transformers) is helpful.
• A basic understanding of machine learning and deep learning concepts with some practical experience.

Recommended Background: If you are new to machine learning, consider attending the preceding tutorials on Machine Learning with Python (Scikit-Learn), Deep Learning with PyTorch, and Open Source Generative AI with Hugging Face to build a solid foundation.

• Researchers, engineers, and data scientists looking to apply LLMs to real document-based tasks.
• Practitioners interested in building RAG-based chatbots for knowledge bases, documentation, or reports.
• Participants with diverse backgrounds (computer science, data science, applied research, domain sciences) who want a practical entry point into Generative AI.

• Those seeking a deep dive into training LLMs from scratch or into the mathematical foundation of NLP or LLMs or optimization theory
• Participants with no prior Python programming experience.

Get ready to explore LLMs, experiment with open-source models, and build your own AI-powered ChatBot!