Sarah M. Olshan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Sarah M. Olshan has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Corey James Richier
Department of Psychology, University of Illinois at Urbana-Champaign, Urbana, IL, USA.
2 shared publications - Kyle A. Baacke
University of Wisconsin-Milwaukee
2 shared publications Wendy Heller
Department of Psychology, University of Illinois Urbana-Champaign
2 shared publications- Christine Anne Vitiello1 shared publication
Corey J. Richier
Department of Psychology, University of Illinois Urbana-Champaign
1 shared publicationGregory A. Miller
Department of Psychology, University of Illinois Urbana-Champaign
1 shared publication
Latest Publications
- Open Access
Comparison of dimensionality reduction and feature selection for cognitive task decoding in functional magnetic resonance imaging
Background: Advances in functional magnetic resonance imaging (fMRI) have led to the ability to study the brain across many contexts. However, the large number of features generated by functional connectivity approaches may overfit the data. These problems can be overcome with either feature selection (FS) or dimensionality reduction (DR), which can be applied to less complex models. We utilize two open-source datasets to compare the performance of DR/FS methods on cognitive task decoding using a suite of ML classifiers. New Method: While DR and FS methods have been used previously in decoding research, no systematic comparison of their performance has been undertaken. Here, we compare available methods using commonly utilized machine learning libraries to establish which methods provide the best predictive performance. We then conduct statistical tests to examine the relative contributions of DR and FS methods and classifiers on decoding accuracy. Results: Neither DR or FS was found to be superior. However, differences were identified across datasets and tasks. In the majority of methods and datasets, a peak in predictive performance was found using a small percentage of the total number of original features. Comparison with existing methods: Some methods perform better than the baseline method of prediction with all available features or selecting features randomly. Decoding performance utilizing the HCP datasets with certain DR/FS methods exceeds that of deep learning approaches. Conclusions: Simple machine learning models with DR/FS have competitive decoding performance. These results suggest a “sweet spot” for the tradeoff between the retention of features and predictive accuracy.
Open AccessComparison of dimensionality reduction and feature selection for cognitive task decoding in functional magnetic resonance imaging
Background: Advances in functional magnetic resonance imaging (fMRI) have led to the ability to study the brain across many contexts. However, the large number of features generated by functional connectivity approaches may overfit the data. These problems can be overcome with either feature selection (FS) or dimensionality reduction (DR), which can be applied to less complex models. We utilize two open-source datasets to compare the performance of DR/FS methods on cognitive task decoding using a suite of ML classifiers. New Method: While DR and FS methods have been used previously in decoding research, no systematic comparison of their performance has been undertaken. Here, we compare available methods using commonly utilized machine learning libraries to establish which methods provide the best predictive performance. We then conduct statistical tests to examine the relative contributions of DR and FS methods and classifiers on decoding accuracy. Results: Neither DR or FS was found to be superior. However, differences were identified across datasets and tasks. In the majority of methods and datasets, a peak in predictive performance was found using a small percentage of the total number of original features. Comparison with existing methods: Some methods perform better than the baseline method of prediction with all available features or selecting features randomly. Decoding performance utilizing the HCP datasets with certain DR/FS methods exceeds that of deep learning approaches. Conclusions: Simple machine learning models with DR/FS have competitive decoding performance. These results suggest a “sweet spot” for the tradeoff between the retention of features and predictive accuracy.
- 1 November 2024
- Published by American Psychological Association (APA) in Journal of Psychopathology and Clinical Science
- Vol. 133 (8) , 678-689
Making the most of errors: Utilizing erroneous classifications generated by machine-learning models of neuroimaging data to capture disorder heterogeneity.
Within-disorder heterogeneity complicates mapping the neurobiological features of psychopathology to Diagnostic and Statistical Manual of Mental Disorders conceptualizations. The present study explored the patterns of diagnostic classification errors among disorders with commonly co-occurring features to examine this heterogeneity. Classification analyses were conducted with the University of California, Los Angeles Phenomics Study database using a support-vector classifier to differentiate disorders via whole brain task-based functional connectivity, predicting that model misclassifications would provide insight about brain connectivity characteristics shared across disorders. Whether symptoms and specific brain networks could account for misclassification rates was also explored. The classification model performed better than chance (44% accuracy, p = .01) and revealed that misclassification of schizophrenia (SCZ) as bipolar disorder (BD; 38%) and BD as SCZ (36%) was symmetrical. Attention-deficit/hyperactivity disorder (ADHD) was misclassified as BD at the highest rate (46%) and higher than the converse (17%). SCZ and ADHD were misclassified least (15% SCZ as ADHD and 22% ADHD as SCZ). Considerable variance in misclassification of SCZ as BD (R2 = .83) and BD as SCZ (R2 = .71) could be accounted for by symptoms of both SCZ and BD. Permutation testing revealed disorder- and network-specific effects, with certain networks improving classification accuracy and others hindering it for specific disorders. An approach focused on classification errors replicated known disorder overlap, producing errors in the expected configuration. Further, it identified clinical and neural features within and across diagnostic categories that contribute to disorder misclassification and within-disorder heterogeneity. This approach may facilitate neurobiologically informed phenotypic differentiation within diagnostic groups. (PsycInfo Database Record (c) 2024 APA, all rights reserved).
- 1 November 2024
- Published by American Psychological Association (APA) in Journal of Psychopathology and Clinical Science
- Vol. 133 (8) , 678-689
Making the most of errors: Utilizing erroneous classifications generated by machine-learning models of neuroimaging data to capture disorder heterogeneity.
Within-disorder heterogeneity complicates mapping the neurobiological features of psychopathology to Diagnostic and Statistical Manual of Mental Disorders conceptualizations. The present study explored the patterns of diagnostic classification errors among disorders with commonly co-occurring features to examine this heterogeneity. Classification analyses were conducted with the University of California, Los Angeles Phenomics Study database using a support-vector classifier to differentiate disorders via whole brain task-based functional connectivity, predicting that model misclassifications would provide insight about brain connectivity characteristics shared across disorders. Whether symptoms and specific brain networks could account for misclassification rates was also explored. The classification model performed better than chance (44% accuracy, p = .01) and revealed that misclassification of schizophrenia (SCZ) as bipolar disorder (BD; 38%) and BD as SCZ (36%) was symmetrical. Attention-deficit/hyperactivity disorder (ADHD) was misclassified as BD at the highest rate (46%) and higher than the converse (17%). SCZ and ADHD were misclassified least (15% SCZ as ADHD and 22% ADHD as SCZ). Considerable variance in misclassification of SCZ as BD (R2 = .83) and BD as SCZ (R2 = .71) could be accounted for by symptoms of both SCZ and BD. Permutation testing revealed disorder- and network-specific effects, with certain networks improving classification accuracy and others hindering it for specific disorders. An approach focused on classification errors replicated known disorder overlap, producing errors in the expected configuration. Further, it identified clinical and neural features within and across diagnostic categories that contribute to disorder misclassification and within-disorder heterogeneity. This approach may facilitate neurobiologically informed phenotypic differentiation within diagnostic groups. (PsycInfo Database Record (c) 2024 APA, all rights reserved).
- Open Access
- 9 October 2024
- Published by Center for Open Science in PsyArXiv
Making the most of errors: Utilizing erroneous classifications generated by machine-learning models of neuroimaging data to capture disorder heterogeneity
Within-disorder heterogeneity complicates mapping neurobiological features of psychopathology to DSM conceptualizations. The present study explored patterns of diagnostic classification errors among disorders with commonly co-occurring features to examine this heterogeneity. We conducted classification analyses with the UCLA Phenomics Study database using a support vector classifier to differentiate disorders via whole-brain task-based functional connectivity, predicting that model misclassifications would provide insight about brain connectivity characteristics shared across disorders. We also explored whether symptoms and specific brain networks could account for misclassification rates. The classification model performed better than chance (44% accuracy, p = .01) and revealed that misclassification of schizophrenia (SCZ) as bipolar disorder (BD; 38%) and BD as SCZ (36%) was symmetrical. Attention-deficit/hyperactivity disorder (ADHD) was misclassified as BD at the highest rate (46%) and higher than the converse (17%). SCZ and ADHD were misclassified least (SCZ as ADHD 15%; ADHD as SCZ 22%). Considerable variance in misclassification of SCZ as BD (R2 = .83) and BD as SCZ (R2 = .71) could be accounted for by symptoms of both SCZ and BD. Permutation testing revealed disorder- and network-specific effects, with certain networks improving classification accuracy and others hindering it for specific disorders. An approach focused on classification errors replicated known disorder overlap, producing errors in the expected configuration. Further, it identified clinical and neural features within and across diagnostic categories that contribute to disorder misclassification and within-disorder heterogeneity. This approach may facilitate neurobiologically informed phenotypic differentiation within diagnostic groups.
- 9 October 2024
- Published by Center for Open Science in PsyArXiv
Open AccessMaking the most of errors: Utilizing erroneous classifications generated by machine-learning models of neuroimaging data to capture disorder heterogeneity
Within-disorder heterogeneity complicates mapping neurobiological features of psychopathology to DSM conceptualizations. The present study explored patterns of diagnostic classification errors among disorders with commonly co-occurring features to examine this heterogeneity. We conducted classification analyses with the UCLA Phenomics Study database using a support vector classifier to differentiate disorders via whole-brain task-based functional connectivity, predicting that model misclassifications would provide insight about brain connectivity characteristics shared across disorders. We also explored whether symptoms and specific brain networks could account for misclassification rates. The classification model performed better than chance (44% accuracy, p = .01) and revealed that misclassification of schizophrenia (SCZ) as bipolar disorder (BD; 38%) and BD as SCZ (36%) was symmetrical. Attention-deficit/hyperactivity disorder (ADHD) was misclassified as BD at the highest rate (46%) and higher than the converse (17%). SCZ and ADHD were misclassified least (SCZ as ADHD 15%; ADHD as SCZ 22%). Considerable variance in misclassification of SCZ as BD (R2 = .83) and BD as SCZ (R2 = .71) could be accounted for by symptoms of both SCZ and BD. Permutation testing revealed disorder- and network-specific effects, with certain networks improving classification accuracy and others hindering it for specific disorders. An approach focused on classification errors replicated known disorder overlap, producing errors in the expected configuration. Further, it identified clinical and neural features within and across diagnostic categories that contribute to disorder misclassification and within-disorder heterogeneity. This approach may facilitate neurobiologically informed phenotypic differentiation within diagnostic groups.
- 1 January 2021
- Published by Psi Chi, the International Honor Society in Psychology in Psi Chi Journal of Psychological Research
- Vol. 26 (3) , 336-346
Need for Cognition and Women’s Implicit Associations: Breaking Stereotypes?
People often associate women more with emotions, or affect in general, compared to men (e.g., Barrett & Bliss-Moreau, 2009); however, it is unknown whether some women will have a stronger association between self and affect or others and cognition than other women. We predicted that higher need for cognition (NFC; Cacioppo et al., 1984), or the enjoyment of cognitive processes, would be associated with stronger self-cognition/others-affect implicit associations. We also predicted that women with stronger self-cognition/others-affect associations would be less likely to endorse STEM stereotypes. We also expected a positive relationship between NFC and explicit self-cognition/others-affect associations. To test these predictions, we conducted a study on Project Implicit investigating the relationship between NFC, affect-cognition associations with the self and others, and endorsement of women in STEM stereotypes (Jackson et al., 2014). We found that higher NFC scores were associated with decreased endorsement of stereotypical affect-cognition implicit associations in women, r(280) = −.14, p = .022, 95% CI [−0.25, −0.02]. There was no significant association between self-affect/others-cognition implicit associations and STEM stereotype endorsement, r(278) = −.05, p = .421, 95% CI [−0.17, 0.07]. Importantly, we found that the relationship between NFC and self-affect/others-cognition associations exists when using both an implicit and explicit measure. NFC may increase the likelihood of women making the counter-stereotypical association of themselves with cognition as opposed to affect. It is important to know who endorses these stereotypes, and future studies should continue to examine the trait NFC, affect-cognition associations, and related interest in STEM.
- 1 January 2021
- Published by Psi Chi, the International Honor Society in Psychology in Psi Chi Journal of Psychological Research
- Vol. 26 (3) , 336-346
Need for Cognition and Women’s Implicit Associations: Breaking Stereotypes?
People often associate women more with emotions, or affect in general, compared to men (e.g., Barrett & Bliss-Moreau, 2009); however, it is unknown whether some women will have a stronger association between self and affect or others and cognition than other women. We predicted that higher need for cognition (NFC; Cacioppo et al., 1984), or the enjoyment of cognitive processes, would be associated with stronger self-cognition/others-affect implicit associations. We also predicted that women with stronger self-cognition/others-affect associations would be less likely to endorse STEM stereotypes. We also expected a positive relationship between NFC and explicit self-cognition/others-affect associations. To test these predictions, we conducted a study on Project Implicit investigating the relationship between NFC, affect-cognition associations with the self and others, and endorsement of women in STEM stereotypes (Jackson et al., 2014). We found that higher NFC scores were associated with decreased endorsement of stereotypical affect-cognition implicit associations in women, r(280) = −.14, p = .022, 95% CI [−0.25, −0.02]. There was no significant association between self-affect/others-cognition implicit associations and STEM stereotype endorsement, r(278) = −.05, p = .421, 95% CI [−0.17, 0.07]. Importantly, we found that the relationship between NFC and self-affect/others-cognition associations exists when using both an implicit and explicit measure. NFC may increase the likelihood of women making the counter-stereotypical association of themselves with cognition as opposed to affect. It is important to know who endorses these stereotypes, and future studies should continue to examine the trait NFC, affect-cognition associations, and related interest in STEM.