Towards Explainable Classifiers Using the Counterfactual Approach - Global Explanations for Discovering Bias in Data

AbstractThe paper proposes summarized attribution-based post-hoc explanations for the detection and identification of bias in data. A global explanation is proposed, and a step-by-step framework on how to detect and test bias is introduced. Since removing unwanted bias is often a complicated and tremendous task, it is automatically inserted, instead. Then, the bias is evaluated with the proposed counterfactual approach. The obtained results are validated on a sample skin lesion dataset. Using the proposed method, a number of possible bias-causing artifacts are successfully identified and confirmed in dermoscopy images. In particular, it is confirmed that black frames have a strong influence on Convolutional Neural Network’s prediction: 22% of them changed the prediction from benign to malignant.

Download Full-text

The effect of raters fatigue on scoring EFL writing tasks

Indonesian Journal of Applied Linguistics ◽

10.17509/ijal.v10i1.24956 ◽

2020 ◽

Vol 10 (1) ◽

pp. 1-13

Author(s):

Amir Mahshanian ◽

Mohammadtaghi Shahnazari

Keyword(s):

Correlation Coefficient ◽

Pearson Correlation ◽

Test Bias ◽

Pearson Correlation Coefficient ◽

Efl Writing ◽

Efl Learners ◽

Negative Effect ◽

Spss Software ◽

Post Hoc ◽

The Relationship

Given the importance of testing, in general, and scoring writing tasks in particular, the negative effect of fatigue on human raters is important to investigate. This study aimed to (1) explore the relationship between fatigue and scoring composition tasks written by upper-intermediate EFL learners; and (2) to investigate the discrepancy of the frequency of comments among EFL raters while scoring composition tasks. Four raters were selected, and each given 28 composition tasks to score and comment on. The data were analyzed through SPSS software by running ANOVA, Pearson correlation coefficient, and post-hoc tests. Results suggested that the scores assigned to the first 16 tasks were significantly lower than those assigned to the last 12 tasks and that the last four tasks were scored highest. Based on the results obtained from the questionnaire, the observed diversity is argued to be rooted in raters’ fatigue and result in test bias. Furthermore, findings indicated that the frequency of comments given by the raters on the first 12 essays was significantly higher than those on the last 16 essays (the highest and the lowest frequency of comments were observed in the first four, and the last four scored essays, respectively).

Download Full-text

Effect of strain on ADF-STEM high-resolution images

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100087641 ◽

1991 ◽

Vol 49 ◽

pp. 666-667

Author(s):

M. Kelly ◽

D.M. Bird

Keyword(s):

High Resolution ◽

Strong Influence ◽

Intensity Variation ◽

Dark Field ◽

Atomic Resolution ◽

Strain Fields ◽

High Resolution Image ◽

Annular Dark Field ◽

High Resolution Images ◽

Interesting Alternative

It is well known that strain fields can have a strong influence on the details of HREM images. This, for example, can cause problems in the analysis of edge-on interfaces between lattice mismatched materials. An interesting alternative to conventional HREM imaging has recently been advanced by Pennycook and co-workers where the intensity variation in the annular dark field (ADF) detector is monitored as a STEM probe is scanned across the specimen. It is believed that the observed atomic-resolution contrast is correlated with the intensity of the STEM probe at the atomic sites and the way in which this varies as the probe moves from cell to cell. As well as providing a directly interpretable high-resolution image, there are reasons for believing that ADF-STEM images may be less suseptible to strain than conventional HREM. This is because HREM images arise from the interference of several diffracted beams, each of which is governed by all the excited Bloch waves in the crystal.

Download Full-text

Identification of hepatitis a virus in cell cultures by solid phase immune electron microscopy

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100142815 ◽

1986 ◽

Vol 44 ◽

pp. 238-239

Author(s):

C.D. Humphrey ◽

T.L. Cromeans ◽

E.H. Cook ◽

D.W. Bradley

Keyword(s):

Electron Microscopy ◽

Liquid Phase ◽

Cell Cultures ◽

Rapid Detection ◽

Hepatitis A ◽

Hepatitis A Virus ◽

Solid Phase ◽

Immune Electron Microscopy ◽

Detection And Identification

There is a variety of methods available for the rapid detection and identification of viruses by electron microscopy as described in several reviews. The predominant techniques are classified as direct electron microscopy (DEM), immune electron microscopy (IEM), liquid phase immune electron microscopy (LPIEM) and solid phase immune electron microscopy (SPIEM). Each technique has inherent strengths and weaknesses. However, in recent years, the most progress for identifying viruses has been realized by the utilization of SPIEM.

Download Full-text

Grain separation enhanced magnetic coercivity in Pt/Co multilayers.

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100151027 ◽

1993 ◽

Vol 51 ◽

pp. 1038-1039 ◽

Cited By ~ 1

Author(s):

G.A. Bertero ◽

R. Sinclair

Keyword(s):

Remanent Magnetization ◽

Strong Influence ◽

Uniaxial Anisotropy ◽

Magnetic Coupling ◽

Recording Media ◽

Magnetic Media ◽

Signal To Noise ◽

Out Of Plane ◽

Longitudinal Recording ◽

The Relationship

Pt/Co multilayers displaying perpendicular (out-of-plane) magnetic anisotropy and 100% perpendicular remanent magnetization are strong candidates as magnetic media for the next generation of magneto-optic recording devices. The magnetic coercivity, Hc, and uniaxial anisotropy energy, Ku, are two important materials parameters, among others, in the quest to achieving higher recording densities with acceptable signal to noise ratios (SNR). The relationship between Ku and Hc in these films is not a simple one since features such as grain boundaries, for example, can have a strong influence on Hc but affect Ku only in a secondary manner. In this regard grain boundary separation provides a way to minimize the grain-to-grain magnetic coupling which is known to result in larger coercivities and improved SNR as has been discussed extensively in the literature for conventional longitudinal recording media.We present here results from the deposition of two Pt/Co/Tb multilayers (A and B) which show significant differences in their coercive fields.

Download Full-text

The Link Between Verbal Short-Term Memory and Anomia Treatment Gains

American Journal of Speech-Language Pathology ◽

10.1044/2019_ajslp-18-0176 ◽

2019 ◽

Vol 28 (3) ◽

pp. 1039-1052

Author(s):

Reva M. Zimmerman ◽

JoAnn P. Silkes ◽

Diane L. Kendall ◽

Irene Minkina

Keyword(s):

Short Term Memory ◽

Controlled Trial ◽

Treatment Success ◽

Linear Regression Models ◽

Language Performance ◽

Short Term ◽

Term Memory ◽

Treatment Gains ◽

Post Hoc

Purpose A significant relationship between verbal short-term memory (STM) and language performance in people with aphasia has been found across studies. However, very few studies have examined the predictive value of verbal STM in treatment outcomes. This study aims to determine if verbal STM can be used as a predictor of treatment success. Method Retrospective data from 25 people with aphasia in a larger randomized controlled trial of phonomotor treatment were analyzed. Digit and word spans from immediately pretreatment were run in multiple linear regression models to determine whether they predict magnitude of change from pre- to posttreatment and follow-up naming accuracy. Pretreatment, immediately posttreatment, and 3 months posttreatment digit and word span scores were compared to determine if they changed following a novel treatment approach. Results Verbal STM, as measured by digit and word spans, did not predict magnitude of change in naming accuracy from pre- to posttreatment nor from pretreatment to 3 months posttreatment. Furthermore, digit and word spans did not change from pre- to posttreatment or from pretreatment to 3 months posttreatment in the overall analysis. A post hoc analysis revealed that only the less impaired group showed significant changes in word span scores from pretreatment to 3 months posttreatment. Discussion The results suggest that digit and word spans do not predict treatment gains. In a less severe subsample of participants, digit and word span scores can change following phonomotor treatment; however, the overall results suggest that span scores may not change significantly. The implications of these findings are discussed within the broader purview of theoretical and empirical associations between aphasic language and verbal STM processing.

Download Full-text