About

Michael Fischer is a Professor of Computer Science at Yale University, with a distinguished academic background including a Ph.D. and M.A. from Harvard University and a B.S. from the University of Michigan. His research interests encompass cryptographic protocols and security, the theory of parallel and distributed systems, and discrete algorithms. Fischer is widely recognized for his work on the distributed consensus problem and for developing the 'parallel prefix' algorithm that underpins the 'scan' operation fundamental to many parallel algorithms. He directed one of the first Ph.D. dissertations on secure and verifiable e-voting in the mid-1980s and has developed information-theoretically secure cryptosystems based on random card deals. Currently, he studies trust from an algorithmic perspective with the goal of enabling e-commerce systems to automatically learn and utilize trust relationships.

Research topics

• Artificial Intelligence
• Political Science
• Computer Science
• Medicine
• Information Retrieval
• Computer Security
• Psychology
• Pedagogy
• Public relations
• Internet privacy
• Mathematics education
• Medical education
• Data science

Selected publications

• Privacy-Preserving Data Sharing for Medical Research

  Lecture notes in computer science · 2021

  1st authorCorresponding
  • Computer Science
  • Computer Science
  • Computer Security
  PublisherDOI

• Soft-Skill Development In Co-Curricular Programs: An Evaluation Of A Community College Student Leadership Program

  2020

  1st authorCorresponding
  • Political Science
  • Medical education
  • Pedagogy

  Colleges are vital for the provision of a trained workforce. As of 2017, over 1,100 community colleges were serving about 41% of the undergraduates across the nation. The emphasis on preparing students to enter the labor market has led to colleges developing integrated academic and co-curricular programs. Many professionals in higher education believe that education happens best in a truly integrative learning environment, where both academic and student affairs programs are used to educate the student. The New Hampshire Community College Student Leadership program was developed to help students gain soft-skills, as identified by the National Association of Colleges and Employers. The objective of this summative evaluation was to discover if the NHCC SLP contributes to the soft skill development of student participants. Two research questions guided the study: RQ1: How do participants in the New Hampshire Community Student Leadership Program perceive the program to have helped them develop soft-skills? RQ2: Do students who participated in the New Hampshire Community College Student Leadership Program feel prepared to enter the workforce at the time of their graduation? As a summative program review, no human participants were used for this study. The evaluation highlighted several NHCC SLP strategies that could be used beyond their student leadership program. The research identified the importance of building learning outcomes into co-curricular programs and the impact that experiential learning has on student engagement. The findings support how co-curricular programs can help prepare students for the workforce. The results revealed that students participated in large part because of their intrinsic motivation; the theoretical framework of self-determination theory supported this finding. Additional data is needed to determine the long term benefits of soft-skill development through co-curricular programs. Recommendations include the need to distinguish between extracurricular activities and co-curricular activities. Co-curricular activities should be based on intentional program design and include targeted learning outcomes. Additional research should be conducted on the long term benefit of soft skill development at the collegiate level. Consideration should be given to the future study of co-curricular programs and soft-skill development utilizing the Goal Content Theory.

  PublisherOA PDF

• Incentives Don't Solve Blockchain's Problems

  2019-10-01

  preprintOpen accessSenior author

  People need a motive to use and maintain a system. In many of the systems we use today, financial rewards and punishments provide a key incentive to participate and abide by the rules. From salaries to traffic tickets, financial motives are often closely tied to a system's viability. Distributed systems such as blockchain also need a mechanism to motivate good behavior. A blockchain must encourage users to maintain the system while preventing a minority of these users from colluding and gaining disproportionate control. Many popular public blockchains use monetary incentives to encourage users to participate and behave appropriately. But these same incentive schemes create more problems than they solve. Mining rewards cause centralization in proof of work chains such as Bitcoin. Validator rewards and punishments invite attacks in proof of stake chains. This paper argues why these incentive schemes are detrimental to blockchain. It also considers a range of other systems-some of which incorporate monetary incentives, some of which do not-to confirm that monetary incentives may be neither necessary nor sufficient for good user behavior.

  PublisherOA PDFDOI

• Online Teaching: Myths, Misconceptions, and Cold Hard Realities

  BRC Journal of Advances in Education · 2018-03-15 · 1 citations

  articleOpen accessSenior author

  There are many interesting strategic and policy issues surrounding the adoption of online programs. However, the heart of any online program -and the key to its success -is with the faculty who develop and teach online courses. This paper focuses on the faculty, addressing common myths and misconceptions about online instruction, and offering a case study on developing online teaching skills.

  PublisherOA PDFDOI

• Scalable Bias-Resistant Distributed Randomness

  2017-05-01 · 272 citations

  articleOpen access

  Bias-resistant public randomness is a critical component in many (distributed) protocols. Generating public randomness is hard, however, because active adversaries may behave dishonestly to bias public random choices toward their advantage. Existing solutions do not scale to hundreds or thousands of participants, as is needed in many decentralized systems. We propose two large-scale distributed protocols, RandHound and RandHerd, which provide publicly-verifiable, unpredictable, and unbiasable randomness against Byzantine adversaries. RandHound relies on an untrusted client to divide a set of randomness servers into groups for scalability, and it depends on the pigeonhole principle to ensure output integrity, even for non-random, adversarial group choices. RandHerd implements an efficient, decentralized randomness beacon. RandHerd is structurally similar to a BFT protocol, but uses RandHound in a one-time setup to arrange participants into verifiably unbiased random secret-sharing groups, which then repeatedly produce random output at predefined intervals. Our prototype demonstrates that RandHound and RandHerd achieve good performance across hundreds of participants while retaining a low failure probability by properly selecting protocol parameters, such as a group size and secret-sharing threshold. For example, when sharding 512 nodes into groups of 32, our experiments show that RandHound can produce fresh random output after 240 seconds. RandHerd, after a setup phase of 260 seconds, is able to generate fresh random output in intervals of approximately 6 seconds. For this configuration, both protocols operate at a failure probability of at most 0.08% against a Byzantine adversary.

  PublisherOA PDFDOI

• Scalable Bias-Resistant Distributed Randomness.

  IACR Cryptology ePrint Archive · 2016-01-01 · 11 citations

  preprint

  Bias-resistant public randomness is a critical component in many (distributed) protocols. Generating public randomness is hard, however, because active adversaries may behave dishonestly to bias public random choices toward their advantage. Existing solutions do not scale to hundreds or thousands of participants, as is needed in many decentralized systems. We propose two large-scale distributed protocols, RandHound and RandHerd, which provide publicly-verifiable, unpredictable, and unbiasable randomness against Byzantine adversaries. RandHound relies on an untrusted client to divide a set of randomness servers into groups for scalability, and it depends on the pigeonhole principle to ensure output integrity, even for non-random, adversarial group choices. RandHerd implements an efficient, decentralized randomness beacon. RandHerd is structurally similar to a BFT protocol, but uses RandHound in a one-time setup to arrange participants into verifiably unbiased random secret-sharing groups, which then repeatedly produce random output at predefined intervals. Our prototype demonstrates that RandHound and RandHerd achieve good performance across hundreds of participants while retaining a low failure probability by properly selecting protocol parameters, such as a group size and secret-sharing threshold. For example, when sharding 512 nodes into groups of 32, our experiments show that RandHound can produce fresh random output after 240 seconds. RandHerd, after a setup phase of 260 seconds, is able to generate fresh random output in intervals of approximately 6 seconds. For this configuration, both protocols operate at a failure probability of at most 0.08% against a Byzantine adversary.

  Publisher

• Private Eyes: Secure Remote Biometric Authentication

  2015-01-01 · 6 citations

  articleOpen access

  We propose an efficient remote biometric authentication protocol that gives strong protection to the user’s biometric data in case of two common kinds of security breaches: (1) loss or theft of the user’s token (smart card, handheld device, etc.), giving the attacker full access to any secrets embedded within it; (2) total penetration of the server. Only if both client and server are simultaneously compromised is the user’s biometric data vulnerable to exposure. The protocol works by encrypting the user’s biometric template in a way that allows it to be used for authentication without being decrypted by either token or server. Further, the encrypted template never leaves the token, and only the server has the information that would enable it to be decrypted. We have implemented our protocol using two iris recognition libraries and evaluated its performance. The overall efficiency and recognition performance is essentially the same compared to an unprotected biometric system.

  PublisherDOI

• Homebuyer Neighborhood Attainment in Black and White: Housing Outcomes during the Housing Boom and Bust

  Social Forces · 2014-12-13 · 26 citations

  article1st authorCorresponding

  This paper examines the types of neighborhoods that black and white homebuyers have secured loans in during the recent housing boom and subsequent bust. We expand upon and refine previous research on locational attainment using loan-level data from the 1992–2010 Home Mortgage Disclosure Act (HMDA) combined with tract- and metropolitan-level data from the 1990, 2000, and 2010 Census and the American Community Survey. Multilevel models show that black homebuyers are moving into considerably more racially segregated neighborhoods than their white counterparts and that their access to “whiter” neighborhoods did not expand during the housing boom, even after controlling for the racial composition of the metropolitan area and other key ecological factors. Conversely, new white homebuyers have been moving into neighborhoods with greater percent black residents, which may be a contributing factor in observed declines in segregation during the past few decades. Additionally, black homebuyers in metropolitan areas with greater suburban growth were on average accessing homes in more integrated neighborhoods. Finally, the models explained considerably more of the variation in the neighborhood racial composition of whites compared to blacks. These findings are suggestive of a dual housing market, one in which the experiences of blacks are still systematically different than those of whites, despite expanded access to homeownership.

  PublisherDOI

• Secure sealed-bid online auctions using discreet cryptographic proofs

  Mathematical and Computer Modelling · 2011-08-03 · 21 citations

  article
  PublisherDOI

• Assigning tasks for efficiency in Hadoop

  2010-06-13 · 54 citations

  article1st authorCorresponding

  In recent years Google’s MapReduce has emerged as a leading large-scale data processing architecture. Adopted by companies such as Amazon, Facebook, Google, IBM and Yahoo! in daily use, and more recently put in use by several universities, it allows parallel processing of huge volumes of data over cluster of machines. Hadoop is a free Java implementation of MapReduce. In Hadoop, files are split into blocks and replicated and spread over all servers in a network. Each job is also split into many small pieces called tasks. Several tasks are processed on a single server, and a job is not completed until all the assigned tasks are finished. A crucial factor that affects the completion time of a job is the particular assignment of tasks to servers. Given a placement of the input data over servers, one wishes to find the assignment that minimizes the total completion time. In this paper, an idealized Hadoop model is proposed to investigate the Hadoop task assignment problem. It is shown that there is no feasible algorithm to find the optimal Hadoop task assignment unless P = NP. Assignments that are computed by the round robin algorithm inspired by the current Hadoop scheduler are shown to deviate from optimum by a multiplicative factor in the worst case. A flow-based algorithm is presented that computes assignments that are optimal to within an additive constant.

  PublisherDOI

Frequent coauthors

• Nancy Lynch

  27 shared

• Albert R. Meyer

  21 shared

• Rebecca N. Wright

  7 shared

• Michael Merritt

  Shell (Netherlands)

  5 shared

• René Peralta

  National Institute of Standards and Technology

  5 shared

• Dike Ahanotu

  4 shared

• Dana Angluin

  4 shared

• Shlomo Moran

  3 shared

Awards & honors

• ACM Fellow