Today's Web-enabled deluge of electronic data calls for automated methods of data analysis. Machine learning provides these, developing methods that can automatically detect patterns in data and then use the uncovered patterns to predict future data. This textbook offers a comprehensive and self-contained introduction to the field of machine learning, based on a unified, probabilistic approach. The coverage combines breadth and depth, offering necessary background material on such topics as probability, optimization, and linear algebra as well as discussion of recent developments in the field, including conditional random fields, L1 regularization, and deep learning. The book is written in an informal, accessible style, complete with pseudo-code for the most important algorithms. All topics are copiously illustrated with color images and worked examples drawn from such application domains as biology, text processing, computer vision, and robotics. Rather than providing a cookbook of different heuristic methods, the book stresses a principled model-based approach, often using the language of graphical models to specify models in a concise and intuitive way. Almost all the models described have been implemented in a MATLAB software package--PMTK (probabilistic modeling toolkit)--that is freely available online. The book is suitable for upper-level undergraduates with an introductory-level college math background and beginning graduate students.

Machine Learning : A Probabilistic Perspective

Bayesian methods have garnered huge interest in cognitive science as an approach to models of cognition and perception. On the other hand, Bayesian methods for data analysis have not yet made much headway in cognitive science against the institutionalized inertia of 20th century null hypothesis significance testing (NHST). Ironically, specific Bayesian models of cognition and perception may not long endure the ravages of empirical verification, but generic Bayesian methods for data analysis will eventually dominate. It is time that Bayesian data analysis became the norm for empirical methods in cognitive science. This article reviews a fatal flaw of NHST and introduces the reader to some benefits of Bayesian data analysis. The article presents illustrative examples of multiple comparisons in Bayesian analysis of variance and Bayesian approaches to statistical power. Copyright © 2010 John Wiley & Sons, Ltd. For further resources related to this article, please visit the WIREs website.

Bayesian data analysis.

Cobb’s concern was a long-worrisome circularity in the sociology of science based on the use of bright lines such as p< 0.05: “We teach it because it’s what we do; we do it because it’s what we teach.” This concern was brought to the attention of the ASA Board. The ASA Board was also stimulated by highly visible discussions over the last few years. For example, ScienceNews (Siegfried 2010) wrote: “It’s science’s dirtiest secret: The ‘scientific method’ of testing hypotheses by statistical analysis stands on a flimsy foundation.” A November 2013, article in Phys.org Science News Wire (2013) cited “numerous deep flaws” in null hypothesis significance testing. A ScienceNews article (Siegfried 2014) on February 7, 2014, said “statistical techniques for testing hypotheses...havemore flaws than Facebook’s privacy policies.” Aweek later, statistician and “Simply Statistics” blogger Jeff Leek responded. “The problem is not that people use P-values poorly,” Leek wrote, “it is that the vast majority of data analysis is not performed by people properly trained to perform data analysis” (Leek 2014). That same week, statistician and science writer Regina Nuzzo published an article in Nature entitled “Scientific Method: Statistical Errors” (Nuzzo 2014). That article is nowone of the most highly viewedNature articles, as reported by altmetric.com (http://www.altmetric.com/details/2115792#score). Of course, it was not simply a matter of responding to some articles in print. The statistical community has been deeply concerned about issues of reproducibility and replicability of scientific conclusions. Without getting into definitions and distinctions of these terms, we observe that much confusion and even doubt about the validity of science is arising. Such doubt can lead to radical choices, such as the one taken by the editors of Basic andApplied Social Psychology, who decided to ban p-values (null hypothesis significance testing) (Trafimow and Marks 2015). Misunderstanding or misuse of statistical inference is only one cause of the “reproducibility crisis” (Peng 2015), but to our community, it is an important one. When the ASA Board decided to take up the challenge of developing a policy statement on p-values and statistical significance, it did so recognizing this was not a lightly taken step. The ASA has not previously taken positions on specific matters of statistical practice. The closest the association has come to this is a statement on the use of value-added models (VAM) for educational assessment (Morganstein and Wasserstein 2014) and a statement on risk-limiting post-election audits (American Statistical Association 2010). However, these were truly policy-related statements. The VAM statement addressed a key educational policy issue, acknowledging the complexity of the issues involved, citing limitations of VAMs as effective performance models, and urging that they be developed and interpreted with the involvement of statisticians. The statement on election auditing was also in response to a major but specific policy issue (close elections in 2008), and said that statistically based election audits should become a routine part of election processes. By contrast, the Board envisioned that the ASA statement on p-values and statistical significance would shed light on an aspect of our field that is too often misunderstood and misused in the broader research community, and, in the process, provides the community a service. The intended audience would be researchers, practitioners, and science writers who are not primarily statisticians. Thus, this statementwould be quite different from anything previously attempted. The Board tasked Wasserstein with assembling a group of experts representing a wide variety of points of view. On behalf of the Board, he reached out to more than two dozen such people, all of whom said theywould be happy to be involved. Several expressed doubt about whether agreement could be reached, but those who did said, in effect, that if there was going to be a discussion, they wanted to be involved. Over the course of many months, group members discussed what format the statement should take, tried to more concretely visualize the audience for the statement, and began to find points of agreement. That turned out to be relatively easy to do, but it was just as easy to find points of intense disagreement. The time came for the group to sit down together to hash out these points, and so in October 2015, 20 members of the group met at the ASA Office in Alexandria, Virginia. The 2-day meeting was facilitated by Regina Nuzzo, and by the end of the meeting, a good set of points around which the statement could be built was developed. The next 3 months saw multiple drafts of the statement, reviewed by group members, by Board members (in a lengthy discussion at the November 2015 ASA Board meeting), and by members of the target audience. Finally, on January 29, 2016, the Executive Committee of the ASA approved the statement. The statement development process was lengthier and more controversial than anticipated. For example, there was considerable discussion about how best to address the issue of multiple potential comparisons (Gelman and Loken 2014). We debated at some length the issues behind the words “a p-value near 0.05 taken by itself offers only weak evidence against the null

https://www.tandfonline.com/doi/pdf/10.1080/00031305.2016.1154108

The ASA's Statement on p-Values: Context, Process, and Purpose

The analysis of medical images has been woven into the fabric of the pattern analysis and machine intelligence (PAMI) community since the earliest days of these Transactions. Initially, the efforts in this area were seen as applying pattern analysis and computer vision techniques to another interesting dataset. However, over the last two to three decades, the unique nature of the problems presented within this area of study have led to the development of a new discipline in its own right. Examples of these include: the types of image information that are acquired, the fully three-dimensional image data, the nonrigid nature of object motion and deformation, and the statistical variation of both the underlying normal and abnormal ground truth. In this paper, we look at progress in the field over the last 20 years and suggest some of the challenges that remain for the years to come.

https://hal.inria.fr/inria-00615100/document

Medical image analysis: progress over two decades and the challenges ahead

(1998). Human cognitive abilities: A survey of factor analytic studies. Gifted and Talented International: Vol. 13, No. 2, pp. 97-98.

Human cognitive abilities: A survey of factor analytic studies

Purpose: To study the angular dependence of optically stimulated luminescencedosimeters (OSLD) in the Radiological Physics Center anthropomorphic quality assurance pelvic phantom to provide accurate dosimetric measurements as a replacement for TLD.Methods: A spherical phantom was constructed to investigate the angular response of the OSLD as oriented in the RPC pelvic phantom. Three OSLD per irradiation angle, placed at the center of the spherical phantom, were irradiated with 100 cGy from six different angles. The angular response at each angle was determined relative to the OSLD response when the beam was incident normally on the OSLD surface. A pelvic phantom dosimetry insert was modified to include both TLD and OSLD. Three treatment plans were developed in Pinnacle v9.0 and one in Accuray's Multiplan, each with increasing angular beam delivery (4 field, IMRT, SmartArc, CyberKnife) for the pelvic phantom using a common dose prescription and constraints. Each plan was delivered to the phantom three times, containing two TLD and two OSLD, oriented in the transverse plane, at the center of the PTV. The dose delivered to the TLD and OSLD was calculated for each treatment and then compared. Results: The angular dependence correction factor for the spherical phantom was found to be uniformly 1.041 ± 0.003 from single beam edge‐on irradiations. The angular dependence correction in the pelvic phantom from multiple beam orientation irradiations was 1.024 ± 0.002, such that the OSLD dose agreed with the TLDdose. Applying the OSLD pelvic phantom correction factor, the RPC measured dose to planning system calculated dose ratio was 0.995 ± 0.009. The established RPC phantom TLDdose to calculated dose ratio was 0.995 ± 0.010. Conclusions: An anthropomorphic phantom OSLD angular dependence correction factor was established such that the final OSLD dose measurements agreed with RPC's TLDdose measurements to within 1%. Work supported by grant CA 10953, awarded by NCI, DHHS.

SU-E-T-86: Development and Implementation of the Use of Optically Stimulated Luminescent Detectors in the Radiological Physics Center Anthropomorphic Quality Assurance Phantoms.

This article extends Bayarri and Berger’s (1999) proposal for model evaluation using “partial posterior” p values to the evaluation of second-stage model assumptions in hierarchical models. Applications focus on normal-normal hierarchical models, although the final example involves an application to a beta-binomial model in which the distribution of the test statistic is assumed to be approximately normal. The notion of using partial posterior p values is potentially appealing because it avoids what the authors refer to as “double use” of the data, that is, use of the data for both fitting model parameters and evaluating model fit. In classical terms, this phenomenon is synonymous to masking and is widely known to reduce the power of test statistics for diagnosing model inadequacy. In the present context, masking is avoided by defining the reference distribution of a test statistic t by the partial posterior distribution, defined as

Comment: Bayesian Checking of the Second Levels of Hierarchical Models

Statistical estimation techniques for medical images require well-founded models for the data formation and acquisition process. In our work, we attempt to quantify the appropriate empirical noise model for two alternative digital chest radiography modalities. In this study, we acquired image sets from the same clinical x-ray unit using photostimulable phosphor plates (PPP) and film/screen. The images were acquired in the range of 1-9 mR. The PPP data, at 1760 by 2140 pixels, were transferred directly to our workstations for analysis, and the films were digitized at 2048 by 2480 pixels (0.2 mm resolution for both modalities). 1024 by 1024 regions of interest (ROIs) without edge effects or artifacts were then extracted. Using Fourier transform methods, the noise power spectra were formed for all the images with smoothing applied near the axes to exclude frequency components due to beam profile, detector inhomogeneity and other structural noise. The noise power spectra were then integrated to find the variance at the exposure in question. This corresponds to an estimate of the non-structural variance. The data for all images obtained with one method were finally combined to form the variance function (variance as a function of the mean) for the modality. The data show that for the range of exposure considered, the PPP and digitized film modalities exhibit significant differences in noise structure. Subtle differences in the noise power spectra were seen. Of particular interest to our application, however, the variance function in the two cases was seen to differ in its functional form. The data from the photostimulable phosphor plates correspond well to a linear model (variance proportional to the mean) throughout most of the range investigated, whereas the digitized film variance function corresponds better to a higher-degree polynomial fit over the range considered. This study indicates that separate models may be warranted for different sources of digital chest radiographs and also suggests the type of assumption that, based on empirical findings, may be useful in such a framework.

Novel statistical analysis of two sources of digital chest radiographs

We describe a Bayesian model for the reconstruction of images based on projection data. The model incorporates a boundary process to sever correlations between neighboring regions within images, and the prior distribution of the boundary process can be modified easily in situations in which precise boundary information is available a priori. An example of a positron emission tomography image reconstructed using boundaries obtained from a high resolution magnetic resonance image is provided.

Image Reconstruction Using A Priori Boundary Information

Surface definition, a process of defining three dimensional surface from volume data, is essential in three dimensional volume data rendering. The traditional method applies a three dimensional gradient operator to the volume data to estimate the strength and orientation of surface present. Applying this method to ultrasound volume data does not produce satisfactory results due to noisy nature of the images and the sensitivity of certain signals to the direction of insonation. A Bayesian approach is proPosed here for surface definition of noisy images in general. We formulate the problem as the estimation of posterior means and standard deviations of Gibbs distributions for surface believability and normal direction. The prior distribution reflects shape properties at multiple scales. The design and implementation of such an approach and its application on ultrasound images are the subject of this paper.

/pdf/surface-estimation-in-ultrasound-images-3hddv94gi8.pdf

Valen E. Johnson

Papers

SU-E-T-86: Development and Implementation of the Use of Optically Stimulated Luminescent Detectors in the Radiological Physics Center Anthropomorphic Quality Assurance Phantoms.

Comment: Bayesian Checking of the Second Levels of Hierarchical Models

Novel statistical analysis of two sources of digital chest radiographs

Image Reconstruction Using A Priori Boundary Information

Surface Estimation in Ultrasound Images