Ph.D. in Data Science
The Ph.D. in Data Science is an interdisciplinary program managed jointly by the School of Engineering and Sciences and the School of Business. The program prepares students for research careers in academia or industry that involve the use of methods and systems for extracting insights from rich data sets, especially as applied to the fields of finances and the life sciences. The program responds to the demand by industry for data scientists with a deep knowledge of the theories, techniques and applications associated with “Big Data” and artificial intelligence. The program also recognizes the broad range of skills needed to successfully apply the tools of the digital revolution in industry. This is reflected in the four core areas of (1) mathematical and statistical modeling, (2) machine learning and artificial intelligence, (3) computational systems, and (4) data management at scale, all of which provide a strong foundation for a thorough strong understanding of (5) a field of application.
Programs of study in two application areas, Financial Services and Life Sciences, are described below. Students may design a program of study in another field of application with support of their advisor and approval of the department chair/program director.
To make progress on leading-edge subjects in a fast moving field like data science requires full-time study. Accordingly, students will be admitted only for full-time on-campus study in partnership with a full-time faculty advisor.
Admission Requirements. The Ph.D. in Data Science is primarily designed for students with technical backgrounds. e.g., an undergraduate or master’s degrees in computer science, computer engineering, business analytics, science or engineering from Stevens or other universities. Applicants to the program must fulfill the following requirements:
A 4-year undergraduate degree from an accredited college or university.
International students for whom English is a second language must demonstrate English language proficiencyby submitting the results of a TOEFL or an IELTS test.
GMAT or GRE test scores not older than 5 years.
Admissions decisions are made beginning in February for the following fall semester. Students are encouraged to apply at any time during the year but it is preferred that complete applications are submitted by January 31.
Credit Requirements. The Ph.D. in Data Science requires 84 credits beyond the bachelor’s degree. A prior master’s degree may be transferred for up to 30 credits without specific course descriptions. The remaining 54 credits must include at least 12 credits of core courses, a minimum of 9 credits of field-specific courses and a minimum of 15 dissertation credits. Approval to enter the Ph.D. in Data Science is generally only given when a student has completed work equivalent to a master’s degree.
Structure of the Ph.D. Program in Data Science
Course work
All courses are worth 3 credits unless otherwise specified.
A minimum of 1 course and a maximum of 3 courses in each of the four core areas (mathematical modeling, machine learning and artificial intelligence, computational systems, and data management at scale).
Completion of the signature doctoral course PRV 961 (3 credits, required of all doctoral students at Stevens) and MGT 719 Research Design (3 credits).
A minimum of three courses (9 credits) in a field of application (e.g., financial services or life sciences).
Doctoral students are expected to maintain a 3.7/4.0 cumulative grade-point average. Students failing to meet this requirement may be placed on probation at the discretion of the faculty.
Scholarly work
Research seminars. Ph.D. students are required to attend research seminars. Students failing to meet this requirement may be put on probation at the discretion of the faculty.
Qualifying Exam. The qualifying exam is an oral examination on a syllabus consisting of research papers, prepared jointly by the student and a committee including the advisor and two tenure-track faculty members. The goal is to establish scholarship in an area of research. The exam needs to be completed by the end of the 4th semester in the program. It consists of a presentation, followed by open-door questions from the audience and a closed-door examination from the committee. The committee can pass, fail, or request re-examination (either written or oral).
Dissertation Work. Students must complete a minimum of 15 credits of DS 960 Dissertation/Research. The dissertation must demonstrate the student’s mastery of the associated topic area, it must exhibit sound research methodology and it must make a unique and substantial contribution to an area of data science research.
Thesis Proposal. By the end of their fourth semester, students must write and present a thesis proposal, where they lay out an intended course of research for their dissertation. The proposal should contain an explanation of the problem and why it is important, a sketch of the proposed solution, and background information that serves to indicate that the problem is unsolved and what prior or related approaches to this or similar problems have already been investigated.
Ph.D. Dissertation Defense. The final Ph.D. dissertation is usually defended at the end of the fourth year of full-time study.
Exceptions. The faculty reserve the right to make exceptions to any of the rules and procedures described above in order to promote and preserve the health of the doctoral program and to ensure each student’s prompt and effective progress through the program.
Prerequisites
A number of prerequisites are expected to be satisfied by the student’s prior undergraduate or master’s degrees before entering the program:
Calculus (2 semesters)
Statistics (1 semester)
Probability (1 semester)
Linear algebra (1 semester)
Fluency in a programming language such as C++ or Java (2 semesters)
Database management (1 semester)
These prerequisites could, for example, be satisfied during the students master of science degree by taking courses equivalent to the following Stevens courses: MA 547 Advanced Calculus I, MA 541 Statistical Methods, MA 540 Introduction to Probability Theory, MA 552 Axiomatic Linear Algebra, CS570 Introduction to Programming, Data Structures, and Algorithms, CS590 Algorithms, and CS 561 Database Management Systems.
Core Courses (Minimum of 12 and maximum of 24 credits)
To acquire the breadth of knowledge necessary for successful research in data science, students must complete at least one and at most three courses in each of the four core areas. Students who demonstrate competency in the topics covered by a core course may, with permission of their advisor, waive the core course and take an approved elective in its place. Students are also required to take the doctoral signature course PRV 961 and MGT719 Research Methods.
Mathematical and Statistical Modeling
| BIA 652 | Multivariate Data Analysis I | 3 |
| FE 542 | Time Series with Applications to Finance | 3 |
| MA 661 | Dynamic Programming and Reinforcement Learning | 3 |
Machine Learning and Artificial Intelligence
| BIA 656 | Advanced Data Analytics and Machine Learning | 3 |
| CS 541 | Artificial Intelligence | 3 |
| CS 559 | Machine Learning: Fundamentals and Applications | 3 |
| FE 690 | Machine Learning in Finance | 3 |
Data Management at Scale
| BIA 678 | Big Data Technologies | 3 |
| CS 522 | Mobile Systems and Applications | 3 |
| CS 609 | Data Management and Exploration on the Web | 3 |
Computational Systems
| BIO 668 | Computational Biology | 3 |
| FE 595 | Financial Technology | 3 |
| CS 549 | Distributed Systems and Cloud Computing | 3 |
| CS 600 | Advanced Algorithm Design and Implementation | 3 |
Signature Doctoral Course (3 Credits)
Research Methodology (3 Credits)
Students who demonstrate competency in the subject area of a particular core course may waive the course with the permission of their advisor and the program director.
Application (Major) Area (Minimum of 9 and maximum of 21 credits)
Depending on their major area of study, and with approval of their advisor, students choose at least three courses from either of the following two lists.
Financial services:
| FE 546 | | |
| FE 545 | Design, Patterns and Derivatives Pricing | 3 |
| FE 550 | Data Visualization Application | 3 |
| FE 610 | Stochastic Calculus for Financial Engineers | 3 |
| FE 635 | Financial Enterprise Risk Engineering | 3 |
| FE 680 | Advanced Derivatives | 3 |
| FIN 638 | Corporate Finance | 3 |
| FIN 628 | Derivatives | 3 |
| FE 655 | Systemic Risk and Financial Regulation | 3 |
| FE 670 | Algorithmic Trading Strategies | 3 |
| FE 621 | Computational Methods in Finance | 3 |
| FIN 703 | Microeconomic Theory | 3 |
| FIN 704 | Econometric Theory and Applications | 3 |
| FIN 705 | Asset Pricing Theory and Applications | 3 |
Life sciences
| CH 664 | Computer Methods in Chemistry | 3 |
| CH 760 | Chemoinformatics | 3 |
| CHE 660 | Advanced Process Control | 3 |
| CHE 661 | Design of Control Systems | 3 |
| CPE 610 | Introduction to Bioinformatics Engineering | 3 |
| CPE 686 | Software Tools in Bioinformatics | 3 |
| CS 544 | Health Informatics | 3 |
General Electives
Students who satisfy the minimum requirements for the program may, with approval of their advisor, take elective courses to make up the 54-credit total course requirement.
Available elective courses include:
| BIA 654 | Experimental Design II | 3 |
| BIA 660 | Web Mining | 3 |
| BIA 662 | Cognitive Computing | 3 |
| BIA 672 | Marketing Analytics | 3 |
| BIA 658 | Social Network Analysis | 3 |
| CPE 646 | Pattern Recognition and Classification | 3 |
| CPE 695 | Applied Machine Learning | 3 |
| CS 522 | Mobile Systems and Applications | 3 |
| CS 549 | Distributed Systems and Cloud Computing | 3 |
| CS 582 | Causal Inference | 3 |
| CS 598 | Visual Information Retrieval | 3 |
| CS 600 | Advanced Algorithm Design and Implementation | 3 |
| CS 601 | Algorithmic Complexity | 3 |
| CS 609 | Data Management and Exploration on the Web | 3 |
| CS 677 | Parallel Programming for Many Core Processors | 3 |
| CS 696 | Database Security | 3 |
| FE 541 | Applied Statistics with Applications in Finance | 3 |
| FE 622 | Simulation Methods in Computational Finance and Economics | 3 |
| FE 635 | Financial Enterprise Risk Engineering | 3 |
| FE 641 | Multivariate Statistics and Advanced Time Series in Finance | 3 |
| FE 646 | Optimization Models and Methods in Finance | 3 |
| FE 655 | Systemic Risk and Financial Regulation | 3 |
| FE 670 | Algorithmic Trading Strategies | 3 |
| FE 672 | Advanced Market Structure and HFT Strategies | 3 |
| FE 710 | Applied Stochastic Differential Equations | 3 |
| FE 720 | Volatility Surface: Risk & Models | 3 |
| MA 541 | Statistical Methods | 3 |
| MA 611 | Probability | 3 |
| MA 612 | Mathematical Statistics | 3 |
| MA 623 | Stochastic Processes | 3 |
| MA 629 | Nonlinear Optimization | 3 |
| MA 630 | Advanced Optimization Methods | 3 |
| MA 641 | Time Series Analysis I | 3 |
| MA 655 | Optimal Control Theory | 3 |
| MA 661 | Dynamic Programming and Reinforcement Learning | 3 |