MS in Applied Data Analytics, AI & Machine Learning Concentration

Prerequisite: Programming experience in any language. Or Instructor's consent. Explore the object-oriented approach to software design and development using Python. You will engage in a detailed discussion of programming concepts starting with the fundamentals of data types, control structures methods, classes, arrays and strings, and proceed to more advanced topics such as inheritance and polymorphism, creating user interfaces, exceptions and streams. Upon completion of this course, you will be able to apply software engineering principles to design and implement Python applications that can be used with analytics and big data. Effective Fall 2021, this course fulfills a single unit in each of the following BU Hub areas: Creativity/Innovation, Critical Thinking, Quantitative Reasoning 2.  [ 4 cr. ]

Prerequisites: MET LB 110 lab and either MET CS520 or MET CS521, or consent of instructor. Learn fundamental components of programs using various data structures to solve computational problems, and implement data structures using a high-level programming language. Algorithms will be created, decomposed, and expressed as pseudocode, and you will analyze their running time and computational complexity.  [ 4 cr. ]

Prerequisites: MET CS 544 or MET CS 550 or consent of instructor. You will learn the foundations of statistical machine learning, regression, and classification, and explore the key components of statistical models, including how to construct, interpret, and evaluate them. Topics include data description and visualization, statistical inference, one- and two-sample tests for means and proportions, simple and multiple linear regression, multinomial and logistic regression, analysis of variance (ANOVA), and regression diagnostics. For each topic, you will examine the methodology, underlying assumptions, interpretation of results, and model assessment. The course includes a programming component using R or Python, providing hands-on experience that reinforces theoretical concepts. Methods are presented through real-world examples to help you understand when and how to apply different statistical techniques effectively.  [ 4 cr. ]

Prerequisite: (MET CS 521) or equivalent or instructor's consent. Students will learn major Python tools and techniques for data analysis. There are weekly assignments and mini projects on topics covered in class. These assignments will help build necessary statistical, visualization and other data science skills for effective use of data science in a variety of applications including finance, text processing, time series analysis and recommendation systems. In addition, students will choose a topic for a final project and present it on the last day of class.  [ 4 cr. ]

Prerequisites: MET LB 103, MET LB 104, and (METCS 520 or METCS 521), or equivalent knowledge, or consent of instructor. The goal of this course is to provide students with the mathematical and practical background required in the field of data analytics. Probability and statistics concepts will be reviewed as well as the R tool for statistical computing and graphics. Different types of data are investigated along with data summarization techniques and plotting methods. Data populations using discrete, continuous, and multivariate distributions are explored. Sampling methods and errors during measurements and computations are analyzed in the course. String manipulations and data wrangling methods are examined in detail. The concepts covered in the course are demonstrated using R. Laboratory Course. Restrictions: This course may not be taken in conjunction with MET CS 550.  [ 4 cr. ]

Prerequisites: Basic knowledge of Python or R; or consent of instructor. - Mathematics is fundamental to data science and machine learning. In this course, you will review essential mathematical concepts and fundamental procedures illustrated by Python and/or R code and visualizations. Computational methods for data science presented through accessible, self-contained examples, intuitive explanations, and visualization will be discussed. Equal emphasis will be placed on both mathematics and computational methods that are at the heart of many algorithms for data analysis and machine learning. You will also advance your mathematical proficiency, enabling you to effectively apply your skills to data analytics and machine learning. Restrictions: This course may not be taken in conjunction with MET CS 544.  [ 4 cr. ]

Prerequisites: MET CS 544, or MET CS 555 or equivalent knowledge, or instructor's consent. - The Web Mining and Graph Analytics course covers the areas of web mining, machine learning fundamentals, text mining, clustering, and graph analytics. This includes learning fundamentals of machine learning algorithms, how to evaluate algorithm performance, feature engineering, content extraction, sentiment analysis, distance metrics, fundamentals of clustering algorithms, how to evaluate clustering performance, and fundamentals of graph analysis algorithms, link analysis and community detection based on graphs. Laboratory Course.  [ 4 cr. ]

Prerequisites: MET CS 521, MET LB 103 and MET LB 104; and either MET CS 579 or MET CS 669; or consent of instructor. - Study basic concepts and techniques of data mining. Topics include data preparation, classification, performance evaluation, association rule mining, regression and clustering. You will learn underlying theories of data mining algorithms in the class and practice those algorithms through assignments and a semester-long class project using R. After finishing this course, you will be able to independently perform data mining tasks to solve real-world problems.  [ 4 cr. ]

Prerequisites: MET CS 248 and MET CS 342. - Study of the ideas and techniques that enable computers to behave intelligently. Search, constraint propagations, and reasoning. Knowledge representation, natural language, learning, question answering, inference, visual perception, and/or problem solving. Laboratory course.  [ 4 cr. ]

Prerequisites: MET CS 767 or consent of instructor. - Investigate reinforcement learning, focusing on fundamental concepts and advanced techniques. You will begin with an introduction to reinforcement learning and key concepts, such as exploitation versus exploration and Markov Decision Processes. Then, as the course progresses, you will delve into state transition diagrams, the Bellman equation, and solutions to the Multi-Armed Bandits problem. Challenges and methods for control and prediction will be explored, as well as tabular methods such as Monte Carlo, Dynamic Programming, Temporal Difference Learning, SARSA, and Q-Learning. The course culminates in a review of neural network concepts, covering convolutional and recurrent neural networks, and approximation methods for both discrete and continuous spaces, including DQN and its variants. Policy gradient methods, actor-critic methods, and ethical considerations in AI and safety issues are also discussed.  [ 4 cr. ]

Prerequisites: MET CS 521 and at least one of MET CS 577, MET CS 622, MET CS 673 or MET CS 682; or consent of instructor. Theories and methods for learning from data. The course covers a variety of approaches, including Supervised and Unsupervised Learning, Regression, k-means, KNN's, Neural Nets and Deep Learning, Transformers, Recurrent Neural Nets, Adversarial Learning, Bayesian Learning, and Genetic Algorithms. The underpinnings are covered: perceptron's, backpropagation, attention, and transformers. Each student creates a term project.  [ 4 cr. ]

Prerequisites: MET CS 577 or consent of instructor. This course provides an in-depth exploration of the critical intersection between Artificial Intelligence (AI) and cybersecurity, focusing on two interconnected themes: protecting AI systems from vulnerabilities and harnessing the power of AI to tackle cybersecurity challenges. As AI becomes a cornerstone of modern technology, ensuring the security of AI-powered systems against adversarial attacks, backdoor attacks, and model theft is essential. Simultaneously, AI offers transformative capabilities for malware detection, intrusion prevention, and malware analysis. Through a combination of theoretical foundations, hands-on exercises, and real-world case studies, students will delve into topics such as adversarial machine learning, backdoor injection and defense, IP protection, and privacy-preserving AI. They will also learn how to design and implement AI-driven tools for identifying and mitigating cyber threats in dynamic environments. The course emphasizes practical applications, encouraging students to build resilient AI systems and utilize advanced AI techniques to enhance system security and detect emerging threats. Hands-on labs based on existing tools are provided and required.  [ 4 cr. ]

Prerequisites: MET CS 577, Python programming, mathematics required for machine learning, and familiarity with neural networks. Or consent of instructor. - The first part of the course covers statistical concepts required for generative artificial intelligence. We review regressions and optimization methods as well as traditional neural network architectures, including perceptron and multilayer perceptron. Next, we move to Convolutional Neural Networks and Recurrent Neural Networks and close this part with Attention and Transformers. The second part of the course focuses on generative neural networks. We start with traditional self-supervised learning algorithms (Self Organized Map and Restricted Boltzmann Machine), then explore Auto Encoder architectures and Generative Adversarial Networks and move toward architectures that construct generative models, including recent advances in NLP, including LLMs, and Retrieval Augmented Methods. Finally, we describe the Neural Radiance Field, 3D Gaussian Splatting, and text-2-image models.  [ 4 cr. ]

Prerequisites: MET CS 566 or instructor's consent. - Students enrolled in this course will gain comprehensive insights into fundamental and advanced concepts within the dynamic realm of computer vision. The curriculum will focus on cutting-edge applications of deep neural networks in computer vision. Through hands-on experiences and practical exercises, students will learn to leverage computer vision and machine learning techniques to solve real-world challenges. This course not only equips students with theoretical knowledge but empowers them to apply these concepts effectively, fostering a deep understanding of how computer vision can be harnessed to address complex problems in diverse industries.  [ 4 cr. ]

This is the first course of the two-part thesis option available to Master’s degree program candidates in the Department of Computer Science. You must have completed at least four courses toward your degree and have a grade point average (GPA) of 3.7 or higher. You are responsible for finding a thesis advisor and a principal reader within the department. Please refer to the Department for further details on the application process. Both MET CS 810 Master’s Thesis 1 and MET CS 811 Master’s Thesis 2 must be completed within 12 months.  [ 4 cr. ]

This is the second course of the two-part thesis option available to Master’s degree program candidates in the Department of Computer Science. You must have completed at least four courses toward your degree and have a grade point average (GPA) of 3.7 or higher. You are responsible for finding a thesis advisor and a principal reader within the department. Please refer to the Department for further details on the application process. Both METCS 810 Master’s Thesis 1 and METCS 811 Master’s Thesis 2 must be completed within 12 months.  [ 4 cr. ]