I310D- Introduction to Human-Centered Data Science is a survey course that introduces students to the theory and practice of data science through a human-centered lens, with emphasis on how design choices influence algorithmic results. Students will gain comfort and facility with fundamental principles of data science including (a) Programming for Data Science with Python (b) Data Engineering (c) Database Systems (d) Machine Learning and (e) Human centered aspects such as privacy, bias, fairness, transparency, accountability, reproducibility, interpretability, and societal implications. Each week’s class divided into two segments: (a) Theory and Methods, a concise description of theoretical concept in data science, and (b) Tutorial, a hands-on session on applying the theory just discussed to a real-world task on publicly available data. We will use Python for programming and cover Python basics in the beginning of the course. For modules related to databases, we will use PostGre SQL.

No description provided.

The class explores the principles of relational database design, and SQL as a query language in depth.

Principles and practices in Data Engineering. Emphasis on the data engineering lifecycle and how to build data pipelines to collect, transform, analyze and visualize data from operational systems. This is a hands-on and highly interactive course. Students will learn analytical data modeling techniques for organizing and querying data. They will learn how to transform data into dimensional models, how to build data products, and how to visualize the data. We will also examine the various roles data engineers can have in an organization and career paths for data professionals

This course will cover relevant fundamental concepts in machine learning (ML) and how they are used to solve real-world problems. Students will learn the theory behind a variety of machine learning tools and practice applying the tools to real-world data such as numerical data, textual data (natural language processing), and visual data (computer vision). Each class is divided into two segments: (a) Theory and Methods, a concise description of an ML concept, and (b) Lab Tutorial, a hands-on session on applying the theory just discussed to a real-world task on publicly available data. We will use Python for programming.   By the end of the course, the goals for the students are to: 1. Develop a sense of where to apply machine learning and where not to, and which ML algorithm to use 2. Understand the process of garnering and preprocessing a variety of “big” real-world data, to be used to train ML systems 3. Characterize the process to train machine learning algorithms and evaluate their performance 4. Develop programming skills to code in Python and use modern ML and scientific computing libraries like SciPy and scikit-learn 5. Propose a novel product/research-focused idea (this will be an iterative process), design and execute experiments, and present the findings and demos to a suitable audience (in this case, the class).

Practical skills and understandings required to effectively work with open source software and understand the projects that build them. Includes git-based collaboration as well as conceptual understanding of licenses, security, technical and social processes in open source development. Class projects involve working with digital trace data from open source repositories.

This course offers students in Information Science a comprehensive exploration into the theories, techniques, and tools of data visualization. It is designed to equip students with the skills to effectively communicate complex information visually, enabling data analysis and decision-making. Through a combination of lectures, hands-on projects, and case studies, students will learn how to design and implement effective and aesthetically appealing data visualizations for a variety of data types and audiences. Upon successful completion of this course, students will be able to: • Understand the principles and psychology of visual perception and how they influence data visualization. • Critically evaluate the effectiveness of different data visualization techniques for varying data types and user needs. • Master the use of leading data visualization tools and libraries such as D3.js, or Tableau. • Develop interactive dashboards and reports that effectively communicate findings to both technical and non-technical audiences. • Apply design principles to create visually appealing, accurate, and accessible data visualizations.

Introduction to the emerging field of Explainable Artificial Intelligence (XAI) from the perspectives of a developer and end-user. Students will gain hands-on experience with some of the most commonly used explainability techniques and algorithms.

Leveraging Text Mining, Natural Language Processing, and Computational Linguistics to address real-world textual data challenges, including document processing, keyword extraction, question answering, translation, summarization, sentiment analysis, search, recommendation, and information extraction. Each week, classes include (a) Theory and Methods for NLP concepts and (b) Lab Tutorials for practical application with Python on multilingual text datasets.

This course lays the foundation for data science education targeting health informatics students interested in learning more broadly about biomedical informatics. No previous coding experience is required. The students will be introduced to basic concepts and tools for data analysis. The focus is on hands-on practice and enjoyable learning. The course will use python as the programming language, and Jupyter Notebooks as the development environment (our “home base”) for the examples, tutorials, and assignments. We use Jupyterlab Notebooks because they are both the industry standard and a nice way to load, visualize, and analyze data and describe our findings in one environment. We will also learn GitHub to document changes and backup our work and, eventually, for use as a collaboration tool. Hands-on data analysis, final projects, and associated presentations will be mandatory for the completion of the course. The outcome for the class is that each student will have a GitHub repository with all of their work (Jupyter notebooks, data, etc.), including a final project that will be presented to the class. Specific topics to be covered include GitHub, Linux/Unix File system, Jupyter Notebooks, Python Programming, and Data Visualization.

This course offers an introduction to Fine-Tuning Open-Source Large Language Models (LLMs) through project-based applications and real-world examples. The course will begin with a foundational understanding of Natural Language Processing (NLP), focusing on Text Preprocessing techniques such as Tokenization and Vectorization. A basic overview of Large Language Models will be provided, covering the fundamental structure and architecture of commonly used Open-Source Frameworks. The course will then focus on three key methods for fine-tuning LLMs: Self-Supervised, Supervised and Reinforcement Learning. Each method will be explored through both theoretical explanations and practical group-based projects, applying these concepts to real-world examples. Students will engage in hands-on projects to strengthen their understanding of how to customize and optimize LLMs for specific tasks or domains.

The class focuses on developing problem solving skills using Python as a programming language. Starting from procedural function development, we also explore object-oriented techniques, and discuss simple data structures that are often used in software development. The students usually do a few programming assignments, take a midterm, and submit a final project.

Explore common data collection, management, and sharing practices in information technology and emerging technologies, such as search engines and AI systems. Students will read papers and engage in discussions about the pros and cons of established data practices and learn about the three main components of responsible data management: 1) consent and ownership, 2) privacy and anonymity, and 3) broader impact. Students will also practice how to collect data, make data-driven decisions, and design data-driven products through group projects as UX designers, researchers, and data scientists. The course will bring in interdisciplinary perspectives with guest speakers from archive science, engineering, and respponsible AI, to provide a holistic view of broader data ecosystems and infrastructures.

This class explores various data science models, both traditional and the state of the art techniques. The course is designed to provide mathematical and computational basis such as Linear Algebra, Optimization techniques, and probabilistic modeling for different types of machine learning models. The goal of the class is provide a foundational basis for data science techniques. The class focuses on PSETs and a final data science project.

Develop prompts for text and image generation through an iterative cycle, making the most of foundation models, including large language models and diffusion models. Overview of the field of prompt engineering, including historical development, ethical dilemmas, and the creation of chatbots.

The current Web has experienced tremendous changes to connect data, people, and knowledge. There are a couple of exciting efforts trying to bring the Web to its full potential. The Semantic Web is one of them which is heavily embedded in the Artificial Intelligence area with the long-term goal to enhance the human and machine interaction by representing data semantics, integrating data silos, and enabling intelligent search and discovery. This course aims to provide the basic overview of the Semantic Web in general, and data semantics in particular, and how they can be applied to enhance data integration and knowledge inference. Ontology is the backbone of the Semantic Web. It models the semantics of data and represents them in markup languages proposed by the World Wide Web Consortium (W3C). W3C plays a significant role in directing major efforts at specifying, developing, and deploying standards for sharing information. Semantically enriched data paves the crucial way to facilitate the Web functionality and interoperability. This course contains three parts: Semantic Web language, RDF graph database (i.e., RDF triple store), and its applications. The fundamental part of the course is the Semantic Web languages. It starts from XML and goes further to RDF and OWL. The RDF graph database part introduces different APIs of Jena and its reasoners. The application part showcases current trends on semantic applications. Prerequisites Basic knowledge of HTML and XML is desired. Course Objectives This course aims to develop a critical appreciation of semantic technologies as they are currently being developed. At the end of this course, students should be able to: • sketch the overall architecture of the Semantic Web. • identify the major technologies of the Semantic Web and explain their roles. • illustrate the design principles of the Semantic Web by applying the technologies. • understand certain limitations of the Semantic Web technologies, and be aware of the kinds of services it can and cannot deliver. Course aims are achieved through: • Lectures covers basic knowledge of the Semantic Web • Projects applying semantic technologies to concrete problems of information delivery and use • Assignments of practicing and utilizing key semantic technologies

This course will cover fundamental concepts in Machine Learning (ML). The course will provide conceptual and practical knowledge on a wide range of modern machine learning algorithms; including supervised learning (multiple linear regression, logistic regression, neural networks, and decision trees), unsupervised learning (clustering, dimensionality reduction, recommender systems), reinforcement learning & deep learning models (CNN, RNN, Autoencoders & Transformers) and also introduce the importance of Prompt Engineering and Retrieval Augmented Generation. The goal is for students to be comfortable and confident in machine learning concepts and have the ablity to build machine learning model solution to challenging real-world problems. If you’re looking to break into AI or build a career in machine learning, this is a great place to start.

Natural Language Processing (NLP) is concerned with interactions between computers and humans through the medium of human languages. It involves analyzing, understanding, and generating human language, making it possible for machines to interpret and respond to human speech and text. NLP is currently making significant contributions to modern technological advancements and serves as the backbone of crucial applications such as Gen AI, Conversational AI, Question Answering, Human Language Translation, Summarization, Sentiment and Emotion Analysis, Search and Recommendation, and Information Extraction in various domains such as healthcare, finance, legal, libraries and education and beyond. The proposed graduate-level course aims to cover fundamental concepts in Natural Language Processing / Computational Linguistics and how they are used to solve real-world problems. Classes in each week will be divided into two segments: (a) Theory and Methods, a concise description of an NLP concept, and (b) Practicum, a hands-on session on applying the theory to a real-world task on publicly available multilingual text datasets. We will use Python for programming along with popular libraries for text processing such as NLTK, SpaCy and HuggingFace's transformers. By the end of the course, the goals for the students are to: 1. Understand the process of garnering and pre-processing a large amount of multilingual textual data from various domains and sources. Characterize the processes to store, load, pre-process multilingual data and apply language processing operations such as normalization, tokenization, lemmatization, chunking and machine readable representation (vector) extraction. 2. Train machine learning algorithms for natural language understanding and generation and evaluate their performance. 3. Learn to extract information from unstructured text and represent them in the form of knowledge graphs 4. Learn to use existing knowledge graphs, ontologies and lexical knowledge networks for predictive analysis on text 5. Learn about popular NLP applications and tasks and the process of building such applications 6. Propose a novel product/research-focused idea (this will be an iterative process), design and execute experiments, and present the findings and demos to a suitable audience (in this case, the class).

Learning key data wrangling maneuvers in abstract and implementations in SQL, Excel, R Tidyverse, and Python Pandas. Maneuvers in data transformations include Nest, Pivot, Mutate (inc. separate/unite), Group/Summarize and Rectangling. Projects include working with "wild caught" data datasets (usually CSV or JSON) and computational notebook environments (e.g., iPython, Jupyter, Rmarkdown, Quarto). Fall 2024 has changes from previous syllabus now that we have Database Design and Introduction to Programming. Nonetheless, the previous syllabus is still useful as it links to course materials that show the teaching approach and type of assignments. http://howisonlab.github.io/datawrangling/#Schedule_of_classes

Introduction to combining human and machine intelligence to benefit people and society. Explore cutting-edge research on a number of subjects related to human-AI interaction, including the psychological and societal impacts of AI as well as design guidelines and methods for human-centered AI.