<div><br></div><div><p>In many complex domains, the input data are often not suited for the typical vector representations used in deep learning models. For example, in knowledge representation, relational learning, and some computer vision tasks, the data are often better represented as graphs or sets. In these cases, a key challenge is to learn a representation function which is invariant to permutations of set or isomorphism of graphs. </p><p>In order to handle graph isomorphism, this thesis proposes a subgraph pattern neural network with invariance to graph isomorphisms and varying local neighborhood sizes. Our key insight is to incorporate the unavoidable dependencies in the training observations of induced subgraphs into both the input features and the model architecture itself via high-order dependencies, which are still able to take node/edge labels into account and facilitate inductive reasoning. </p><p>In order to learn permutation-invariant set functions, this thesis shows how the characteristics of an architecture’s computational graph impact its ability to learn in contexts with complex set dependencies, and demonstrate limitations of current methods with respect to one or more of these complexity dimensions. I also propose a new Self-Attention GRU architecture, with a computation graph that is built automatically via self-attention to minimize average interaction path lengths between set elements in the architecture’s computation graph, in order to effectively capture complex dependencies between set elements.</p><p>Besides the typical set problem, a new problem of representing sets-of-sets (SoS) is proposed. In this problem, multi-level dependence and multi-level permutation invariance need to be handled jointly. To address this, I propose a hierarchical sequence attention framework (HATS) for inductive set-of-sets embeddings, and develop the stochastic optimization and inference methods required for efficient learning.</p></div>

The Santa Clara Brief Compassion Scale (SCBCS) is a brief measure of compassion, created in English and translated into Brazilian Portuguese. Nonetheless, to date, no study has assessed the psychometric evidence of its Spanish translation. This study examines the evidence of validity, reliability, and factorial invariance according to the gender of a Spanish version of the SCBCS. Participants included 273 Peruvian university students (50.9% women) with an average age of 21.23 years (SD = 3.24); divided into two groups of men and women to conduct the invariance factor analysis. Other measures of mindfulness, well-being, empathy, and anxiety were applied along with the SCBCS. The Confirmatory Factor Analysis (CFA) indicated that a unifactorial model adjusted significantly to the data (χ2 = 12,127, df = 5, p =.033, χ2 /df = 2.42, CFI =.998, RMSEA =.072 [CI90%.019,.125]; SRMR =.030, WRMR =.551) and presented good reliability (α =.90 [95%.88–.92]; ω =.91). Moreover, correlations between the SCBCS and other measures of mindfulness (r =.53, p <.05, cognitive empathy (r = 55; p <.05), affective empathy (r =.56, p <.05), well-being (r =.55, p <.05), and anxiety (r = −.46; p <.05) supported the convergent and discriminant validity. Likewise, the multiple-group CFA supported the factorial invariance according to the gender of the SCBCS. Results indicate that the SCBCS possesses evidence of validity, reliability, and invariance between men and women for measuring compassion toward others in Peruvian undergraduate students. SCBCS is expected to be used by researchers, healthcare professionals, teachers, and others as a useful measure of compassion in college students.

To guide action, animals anticipate what events will occur, and when they will occur, based on experience. How animals anticipate future events is an unsettled question. Although reinforcement learning is often used to model anticipation, it is resource-intensive outside of the simplest scenarios. In this dissertation, I show evidence of memory that is persistent and carries timing information, and specify an algorithm for how animals might anticipate the identity and timing of future events. This dissertation consists of two studies. In the first study, I found that identity and timing of remembered odors are jointly represented in the same cells in the dentate gyrus and lateral entorhinal cortex. Further, odor memories persist well after new odors emerge. The study analyzed results from an experiment conducted by Woods et al. (2020) on mice passively exposed to separate odors for a period of 20 s per exposure. The results are consistent with a memory framework known as timing using inverse Laplace transform (TILT). In the second study, I constructed a computational algorithm based on the TILT memory framework to anticipate the identity and timing of future events. The algorithm generates predictions based on memories of past events, and stored associations between cues and outcomes. The algorithm is resource-efficient even when the future depends on the indefinite past. The algorithm is scale-invariant and works well with chains of events. Together, the studies support a novel computational mechanism which anticipates what events will occur, and when they will occur. The algorithm could be applied in machine learning in cases of long-range dependence on history. These studies predict that behavioral and neural responses of animals could depend on events well into the past. / 2024-03-13T00:00:00Z

Indiana University-Purdue University Indianapolis (IUPUI) / Previous research has identified impulsive personality traits as significant risk factors for a wide range of risk-taking behavior, substance use, and clinical problems. Most work has been conducted in primarily White samples, leaving it unclear whether these patterns generalize to racial and ethnic minorities, who have higher rates of negative consequences of substance use behavior. The most widely used assessment of impulsive traits is the UPPS-P Impulsive Behavior scale, which has strong psychometric properties across demographic subgroups, such as gender and age; however, data supporting its use in racial and ethnic minorities is less well-developed. The aims of this study are to 1) examine the measurement invariance of the UPPS-P Impulsive Behavior Scale-Short Form (Cyders et al., 2014) across racial minority groups and 2) determine if impulsive personality traits differentially relate to substance use outcomes across racial groups. Participants were 1301 young adults (ages 18-35, fluent in English), recruited through an online survey for both college students at a large public university and Mechanical Turk, a crowdsourcing online platform. Measurement invariance was assessed using multigroup confirmatory factor analysis. Differential validity was assessed using a structural equation modeling framework. I established model fit for each racial group (White group: RMSEA= .067, CFI= .94; Black group: RMSEA= .071, 90% CFI= .952; Asian American group: RMSEA= .073, CFI= .94; Hispanic group: RMSEA=.081, CFI=.934). Based on change in CFI/RMSEA indices, I concluded strong measurement invariance of the Short UPPS-P as a valid scale of impulsive behavior across racial groups. In the White group, findings indicated significant relationships between multiple SUPPS-P traits and alcohol and substance use. In the Asian American group, positive relationships were found between sensation and alcohol use (p=.015) and negative urgency and drug use (p=.020). I found that there were no differences in the relationships between the Short UPPS-P traits and substance use outcomes across White and the racial and ethnic groups studied (p’s>.06).

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第20152号 / 理博第4237号 / 新制||理||1609(附属図書館) / 京都大学大学院理学研究科数学・数理解析専攻 / (主査)教授 堤 誉志雄, 教授 上田 哲生, 教授 國府 寛司 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DFAM

This study examined the Casey Foster Applicant Inventory-Applicant-Co-Parenting Scale (CFAI-CP), a new scale developed to measure foster parent applicants' co-parenting potential. Also, this study illustrates statistical methods used to analyze the psychometric properties of dyadic data. Factor structure and measurement invariance were tested with 111 approved foster couples. Mplus was used to accommodate ordinal-level data. Exploratory factor analysis supported a 10-item, unidimensional measure with excellent internal consistency reliability (.88 fathers,.89 mothers). Confirmatory factor analysis supported scalar measurement invariance but not structural invariance, as expected. Good construct validity was evident. Findings support the CFAI-CP as an empirically sound measure to assess foster parent co-parenting.

Rasch is a mathematical model that allows researchers to compare data that measure a unidimensional trait or ability (Bond & Fox, 2007). When data fit the Rasch model, it is mathematically proven that the item difficulty estimates are independent of the sample of respondents. The purpose of this study was to test the robustness of the Rasch model with regards to its ability to maintain invariant item difficulty estimates when real (data that does not perfectly fit the Rasch model), polytomous scored data is used. The data used in this study comes from a university alumni questionnaire that was collected over a period of five years. The analysis tests for significant variation between (a) small samples taken from a larger sample, (b) a base sample and subsequent (longitudinal) samples and (c) variation over time with confounding variables. The confounding variables studied include (a) the gender of the respondent and (b) the respondent's type of major at the time of graduation. The study used three methods to assess variation: (a) the between-fit statistic, (b) confidence intervals around the mean of the estimates and (c) a general linear model. The general linear model used the person residual statistic from the Winsteps' person output file as a dependent variable with year, gender and type of major as independent variables. Results of the study support the invariant nature of the item difficulty estimates when polytomous data from the alumni questionnaire is used. The analysis found comparable results (within sampling error) for the between-fit statistics and the general linear model. The confidence interval method was limited in its usefulness due to small confidence bands and the limitation of the plots. The linear model offered the most valuable data in that it provides methods to not only detect the existence of variation but to assess the relative magnitude of the variation from different sources. Recommendations for future research include studies regarding the impact of sample size on the between-fit statistic and confidence intervals as well as the impact of large amounts of systematic missing data on the item parameter estimates.

With increased data sharing and research collaboration options available through modern technology, there is an increased need to find more advanced techniques to analyze data across multiple studies. A systematic method of pooling participant-level versus study-level data would be particularly valuable as it would allow for more complex statistical analyses, broader assessment of constructs, and a cost effective way to examine new questions and replicate previous findings. One notable difficulty in pooling raw data in the behavioral sciences is the heterogeneity in methodologies and consequent need to establish measurement invariance. The present study explores the feasibility of using Integrative Data Analysis (IDA) to combine 10 heterogeneous eating disorder prevention data sets and establish measurement invariance across the constructs of thin ideal internalization and body dissatisfaction. Using standard multiple groups factor analysis and likelihood-ratio tests to examine differential item functioning, separate one-factor models were established for the three measures used across studies. Partial measurement invariance was established for all measures. Implications for future IDA studies based on this process are discussed, particularly regarding the clinical impact of measurement invariance.

The perception and understanding of human motion and action is an important area of research in computer vision that plays a crucial role in various applications such as surveillance, HCI, ergonomics, etc. In this thesis, we focus on the recognition of actions in the case of varying viewpoints and different and unknown camera intrinsic parameters. The challenges to be addressed include perspective distortions, differences in viewpoints, anthropometric variations, and the large degrees of freedom of articulated bodies. In addition, we are interested in methods that require little or no training. The current solutions to action recognition usually assume that there is a huge dataset of actions available so that a classifier can be trained. However, this means that in order to define a new action, the user has to record a number of videos from different viewpoints with varying camera intrinsic parameters and then retrain the classifier, which is not very practical from a development point of view. We propose algorithms that overcome these challenges and require just a few instances of the action from any viewpoint with any intrinsic camera parameters. Our first algorithm is based on the rank constraint on the family of planar homographies associated with triplets of body points. We represent action as a sequence of poses, and decompose the pose into triplets. Therefore, the pose transition is broken down into a set of movement of body point planes. In this way, we transform the non-rigid motion of the body points into a rigid motion of body point iii planes. We use the fact that the family of homographies associated with two identical poses would have rank 4 to gauge similarity of the pose between two subjects, observed by different perspective cameras and from different viewpoints. This method requires only one instance of the action. We then show that it is possible to extend the concept of triplets to line segments. In particular, we establish that if we look at the movement of line segments instead of triplets, we have more redundancy in data thus leading to better results. We demonstrate this concept on “fundamental ratios.” We decompose a human body pose into line segments instead of triplets and look at set of movement of line segments. This method needs only three instances of the action. If a larger dataset is available, we can also apply weighting on line segments for better accuracy. The last method is based on the concept of “Projective Depth”. Given a plane, we can find the relative depth of a point relative to the given plane. We propose three different ways of using “projective depth:” (i) Triplets - the three points of a triplet along with the epipole defines the plane and the movement of points relative to these body planes can be used to recognize actions; (ii) Ground plane - if we are able to extract the ground plane, we can find the “projective depth” of the body points with respect to it. Therefore, the problem of action recognition would translate to curve matching; and (iii) Mirror person - We can use the mirror view of the person to extract mirror symmetric planes. This method also needs only one instance of the action. Extensive experiments are reported on testing view invariance, robustness to noisy localization and occlusions of body points, and action recognition. The experimental results are very promising and demonstrate the efficiency of our proposed invariants. iv

This present study examined the structural validity of a Chinese version of Multidimensional Work Ethic Profile (MWEP-C), using a large sample of Chinese parents and their young adult children (N = 1047). Confirmatory factor analysis (CFA) was applied to evaluate the model fit of sample data on three competing models using two randomly split stratified subsamples. Measurement invariance for these two generational respondents was checked using differential item functioning (DIF) analysis. The results indicated that MWEP-C provided a reasonable fit for the sample data and the majority of survey items produced similar item-level responses for individuals that do not differ on the attributes of work ethic across these two generations. DIF items were detected based on advanced and successive iterations. Monte Carlo simulations were also conducted for creating threshold values and for chi-square probabilities based on 1,000 replications. After identifying the DIF items, model fit improved and generational differences and similarities in work ethic between parents and their young adult children were also identified. The results suggested that the younger Chinese generations have higher work ethic mean scores on the dimensions of work centrality and morality/ethics while they have similarities on time concept, self-reliance, delay of gratification, and hard work as their parents.