Donald Geman

Last updated
Donald J. Geman
DonaldGeman.jpg
Donald Geman (right), Fall 1983, Paris
Born (1943-09-20) September 20, 1943 (age 80)
Chicago, Illinois, United States
Nationality American
Alma mater Columbia University
University of Illinois Urbana-Champaign
Northwestern University
Relatives Stuart Geman (brother)
Awards ISI highly cited researcher
Scientific career
Fields Mathematics
Statistics
Institutions University of Massachusetts
Johns Hopkins University
École Normale Supérieure de Cachan
Doctoral advisor Michael Marcus

Donald Jay Geman (born September 20, 1943) is an American applied mathematician and a leading researcher in the field of machine learning and pattern recognition. He and his brother, Stuart Geman, are very well known for proposing the Gibbs sampler and for the first proof of the convergence of the simulated annealing algorithm, [1] in an article that became a highly cited reference in engineering (over 21K citations according to Google Scholar, as of January 2018). [2] He is a professor at the Johns Hopkins University and simultaneously a visiting professor at École Normale Supérieure de Cachan.

Contents

Biography

Geman was born in Chicago in 1943. He graduated from the University of Illinois Urbana-Champaign in 1965 with a B.A. degree in English Literature and from Northwestern University in 1970 with a Ph.D. in mathematics. [3] His dissertation was entitled as "Horizontal-window conditioning and the zeros of stationary processes." He joined University of Massachusetts - Amherst in 1970, where he retired as a distinguished professor in 2001. Thereafter, he became a professor at the Department of Applied Mathematics at Johns Hopkins University. He has also been a visiting professor at the École Normale Supérieure de Cachan since 2001. He is a member of the National Academy of Sciences, and Fellow of the Institute of Mathematical Statistics and the Society for Industrial and Applied Mathematics.

Work

D. Geman and J. Horowitz published a series of papers during the late 1970s on local times and occupation densities of stochastic processes. A survey of this work and other related problems can be found in the Annals of Probability. [4] In 1984 with his brother Stuart, he published a milestone paper which is still today one of the most cited papers [5] in the engineering literature. It introduces a Bayesian paradigm using Markov Random Fields for the analysis of images. This approach has been highly influential over the last 20 years and remains a rare tour de force in this rapidly evolving field. In another milestone paper, [6] [7] in collaboration with Y. Amit, he introduced the notion for randomized decision trees, [8] [9] which have been called random forests and popularized by Leo Breiman. Some of his recent works include the introduction of coarse-to-fine hierarchical cascades for object detection [10] in computer vision and the TSP (Top Scoring Pairs) classifier as a simple and robust rule for classifiers trained on high dimensional small sample datasets in bioinformatics. [11] [12]

Related Research Articles

<span class="mw-page-title-main">Stochastic process</span> Collection of random variables

In probability theory and related fields, a stochastic or random process is a mathematical object usually defined as a sequence of random variables in a probability space, where the index of the sequence often has the interpretation of time. Stochastic processes are widely used as mathematical models of systems and phenomena that appear to vary in a random manner. Examples include the growth of a bacterial population, an electrical current fluctuating due to thermal noise, or the movement of a gas molecule. Stochastic processes have applications in many disciplines such as biology, chemistry, ecology, neuroscience, physics, image processing, signal processing, control theory, information theory, computer science, and telecommunications. Furthermore, seemingly random changes in financial markets have motivated the extensive use of stochastic processes in finance.

In statistics, Markov chain Monte Carlo (MCMC) is a class of algorithms used to draw samples from a probability distribution. Given a probability distribution, one can construct a Markov chain whose elements' distribution approximates it – that is, the Markov chain's equilibrium distribution matches the target distribution. The more steps that are included, the more closely the distribution of the sample matches the actual desired distribution.

Zohar Manna was an Israeli-American computer scientist who was a professor of computer science at Stanford University.

Random forests or random decision forests is an ensemble learning method for classification, regression and other tasks that operates by constructing a multitude of decision trees at training time. For classification tasks, the output of the random forest is the class selected by most trees. For regression tasks, the mean or average prediction of the individual trees is returned. Random decision forests correct for decision trees' habit of overfitting to their training set.

<span class="mw-page-title-main">Yves Meyer</span> French mathematician

Yves F. Meyer is a French mathematician. He is among the progenitors of wavelet theory, having proposed the Meyer wavelet. Meyer was awarded the Abel Prize in 2017.

<span class="mw-page-title-main">Wendelin Werner</span>

Wendelin Werner is a German-born French mathematician working on random processes such as self-avoiding random walks, Brownian motion, Schramm–Loewner evolution, and related theories in probability theory and mathematical physics. In 2006, at the 25th International Congress of Mathematicians in Madrid, Spain he received the Fields Medal "for his contributions to the development of stochastic Loewner evolution, the geometry of two-dimensional Brownian motion, and conformal field theory". He is currently Rouse Ball professor of Mathematics at the University of Cambridge.

<span class="mw-page-title-main">Elisabeth Pate-Cornell</span> American engineer

Marie-Elisabeth Lucienne Paté-Cornell,, is a Stanford University Professor in the Department of Management Science and Engineering and was the Founding Chair of the Department.

Charles "Chip" Lawrence is an American bioinformatician and mathematician, who is the pioneer in developing novel statistical approaches to biological sequence analysis.

<span class="mw-page-title-main">Stuart Geman</span> American mathematician

Stuart Alan Geman is an American mathematician, known for influential contributions to computer vision, statistics, probability theory, machine learning, and the neurosciences. He and his brother, Donald Geman, are well known for proposing the Gibbs sampler, and for the first proof of convergence of the simulated annealing algorithm.

<span class="mw-page-title-main">Raphael Douady</span> French mathematician and economist

Raphael Douady is a French mathematician and economist. He holds the Robert Frey Endowed Chair for Quantitative Finance at Stony Brook, New York. He is a fellow of the Centre d’Economie de la Sorbonne, Paris 1 Pantheon-Sorbonne University, and academic director of the Laboratory of Excellence on Financial Regulation.

<span class="mw-page-title-main">Grégory Miermont</span> French mathematician

Grégory Miermont is a French mathematician working on probability, random trees and random maps.

<span class="mw-page-title-main">École normale supérieure de Rennes</span>

The École normale supérieure de Rennes, also called ENS Rennes is a French scientific grande école, belonging to the network of écoles normales supérieures established according to the model of the École normale supérieure in Paris. Like its sister universities, its mandate lies in training students with a view to careers in academia, engineering and government.

<span class="mw-page-title-main">Gérard Ben Arous</span> French mathematician

Gérard Ben Arous is a French mathematician, specializing in stochastic analysis and its applications to mathematical physics. He served as the director of the Courant Institute of Mathematical Sciences at New York University from 2011 to 2016.

Sandrine Péché is a French mathematician who works as a professor in the Laboratoire de Probabilités, Statistique et Modélisation of Paris Diderot University. Her research concerns probability theory, mathematical physics, and the theory and applications of random matrices.

<span class="mw-page-title-main">Michael I. Miller</span> American biomedical engineer and neuroscientist

Michael Ira Miller is an American-born biomedical engineer and data scientist, and the Bessie Darling Massey Professor and Director of the Johns Hopkins University Department of Biomedical Engineering. He worked with Ulf Grenander in the field of Computational Anatomy as it pertains to neuroscience, specializing in mapping the brain under various states of health and disease by applying data derived from medical imaging. Miller is the director of the Johns Hopkins Center for Imaging Science, Whiting School of Engineering and codirector of Johns Hopkins Kavli Neuroscience Discovery Institute. Miller is also a Johns Hopkins University Gilman Scholar.

Alan Enoch Gelfand is an American statistician, and is currently the James B. Duke Professor of Statistics and Decision Sciences at Duke University. Gelfand’s research includes substantial contributions to the fields of Bayesian statistics, spatial statistics and hierarchical modeling.

The following outline is provided as an overview of and topical guide to machine learning:

In network theory, collective classification is the simultaneous prediction of the labels for multiple objects, where each label is predicted using information about the object's observed features, the observed features and labels of its neighbors, and the unobserved labels of its neighbors. Collective classification problems are defined in terms of networks of random variables, where the network structure determines the relationship between the random variables. Inference is performed on multiple random variables simultaneously, typically by propagating information between nodes in the network to perform approximate inference. Approaches that use collective classification can make use of relational information when performing inference. Examples of collective classification include predicting attributes of individuals in a social network, classifying webpages in the World Wide Web, and inferring the research area of a paper in a scientific publication dataset.

In the domain of physics and probability, the filters, random fields, and maximum entropy (FRAME) model is a Markov random field model of stationary spatial processes, in which the energy function is the sum of translation-invariant potential functions that are one-dimensional non-linear transformations of linear filter responses. The FRAME model was originally developed by Song-Chun Zhu, Ying Nian Wu, and David Mumford for modeling stochastic texture patterns, such as grasses, tree leaves, brick walls, water waves, etc. This model is the maximum entropy distribution that reproduces the observed marginal histograms of responses from a bank of filters, where for each filter tuned to a specific scale and orientation, the marginal histogram is pooled over all the pixels in the image domain. The FRAME model is also proved to be equivalent to the micro-canonical ensemble, which was named the Julesz ensemble. Gibbs sampler is adopted to synthesize texture images by drawing samples from the FRAME model.

<span class="mw-page-title-main">Florent Krzakala</span> French physicist and engineer

Florent Krzakala is a French physicist and applied mathematician, currently a professor at EPFL. His research focuses on the resolution of theoretical problems in physics, computer science, machine learning, statistics and signal processing using mathematical tools inspired by the field of statistical physics.

References

  1. S. Geman; D. Geman (1984). "Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images". IEEE Transactions on Pattern Analysis and Machine Intelligence . 6 (6): 721–741. doi:10.1109/TPAMI.1984.4767596. PMID   22499653. S2CID   5837272.
  2. Google Scholar: Stochastic Relaxation, Gibbs Distributions and the Bayesian Restoration.
  3. "Donald Geman elected to NAS". Institute of Mathematical Statistics. 18 May 2015. Retrieved 5 June 2024.
  4. D. Geman; J. Horowitz (1980). "Occupation Densities". Annals of Probability. 8 (1): 1–67. doi: 10.1214/aop/1176994824 .
  5. ISI Highly Cited:Donald Geman http://hcr3.isiknowledge.com/author.cgi?&link1=Search&link2=Search%20Results&AuthLastName=geman&AuthFirstName=&AuthMiddleName=&AuthMailnstName=&CountryID=-1&DisciplineID=0&id=519 Archived 2007-05-19 at the Wayback Machine
  6. Y. Amit and D. Geman, "Randomized inquiries about shape; an application to handwritten digit recognition," Technical Report 401, Department of Statistics, University of Chicago, IL, 1994.
  7. Y. Amit; D. Geman (1997). "Shape Quantization and Recognition with Randomized Trees". Neural Computation. 9 (7): 1545–1588. CiteSeerX   10.1.1.57.6069 . doi:10.1162/neco.1997.9.7.1545. S2CID   12470146.
  8. Decision Forests: A Unified Framework for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning Found. Trends. Comput. Graph. Vis., Vol. 7, Nos. 2–3 (2011) 81–227. (February 2012), pp. 81-227,doi:10.1561/0600000035 by Antonio Criminisi, Jamie Shotton and Ender Konukoglu.
  9. Decision Forests for Computer Vision and Medical Image Analysis. Editors: A. Criminisi, J. Shotton. Springer, 2013. ISBN   978-1-4471-4928-6 (Print) 978-1-4471-4929-3 (Online).
  10. F. Fleuret; D. Geman (2001). "Coarse-to-Fine Face Detection". International Journal of Computer Vision. 41: 85–107. doi:10.1023/a:1011113216584. S2CID   6754141.
  11. D. Geman; C. d'Avignon; D. Naiman; R. Winslow (2004). "Classifying gene expression profiles from pairwise mRNA comparisons". Statistical Applications in Genetics and Molecular Biology. 3: 1–19. doi:10.2202/1544-6115.1071. PMC   1989150 . PMID   16646797.
  12. A-C Tan; D. Naiman; L. Xu; R. Winslow; D. Geman (2005). "Simple decision rules for classifying human cancers from gene expression profiles". Bioinformatics. 21 (20): 3896–3904. doi:10.1093/bioinformatics/bti631. PMC   1987374 . PMID   16105897.