Search for: All records

There is a growing body of research revealing that longitudinal passive sensing data from smartphones and wearable devices can capture daily behavior signals for human behavior modeling, such as depression detection. Most prior studies build and evaluate machine learning models using data collected from a single population. However, to ensure that a behavior model can work for a larger group of users, its generalizability needs to be verified on multiple datasets from different populations. We present the first work evaluating cross-dataset generalizability of longitudinal behavior models, using depression detection as an application. We collect multiple longitudinal passive mobile sensing datasets with over 500 users from two institutes over a two-year span, leading to four institute-year datasets. Using the datasets, we closely re-implement and evaluated nine prior depression detection algorithms. Our experiment reveals the lack of model generalizability of these methods. We also implement eight recently popular domain generalization algorithms from the machine learning community. Our results indicate that these methods also do not generalize well on our datasets, with barely any advantage over the naive baseline of guessing the majority. We then present two new algorithms with better generalizability. Our new algorithm, Reorder, significantly and consistently outperforms existing methods on most cross-dataset generalization setups. However, the overall advantage is incremental and still has great room for improvement. Our analysis reveals that the individual differences (both within and between populations) may play the most important role in the cross-dataset generalization challenge. Finally, we provide an open-source benchmark platform GLOBEM- short for Generalization of Longitudinal BEhavior Modeling - to consolidate all 19 algorithms. GLOBEM can support researchers in using, developing, and evaluating different longitudinal behavior modeling methods. We call for researchers' attention to model generalizability evaluation for future longitudinal human behavior modeling studies. 

A schizophrenia relapse has severe consequences for a patient’s health, work, and sometimes even life safety. If an oncoming relapse can be predicted on time, for example by detecting early behavioral changes in patients, then interventions could be provided to prevent the relapse. In this work, we investigated a machine learning based schizophrenia relapse prediction model using mobile sensing data to characterize behavioral features. A patient-independent model providing sequential predictions, closely representing the clinical deployment scenario for relapse prediction, was evaluated. The model uses the mobile sensing data from the recent four weeks to predict an oncoming relapse in the next week. We used the behavioral rhythm features extracted from daily templates of mobile sensing data, self-reported symptoms collected via EMA (Ecological Momentary Assessment), and demographics to compare different classifiers for the relapse prediction. Naive Bayes based model gave the best results with an F2 score of 0.083 when evaluated in a dataset consisting of 63 schizophrenia patients, each monitored for up to a year. The obtained F2 score, though low, is better than the baseline performance of random classification (F2 score of 0.02 ± 0.024). Thus, mobile sensing has predictive value for detecting an oncoming relapse and needs further investigation to improve the current performance. Towards that end, further feature engineering and model personalization based on the behavioral idiosyncrasies of a patient could be helpful. 

Abstract Next-generation wearable electronics require enhanced mechanical robustness and device complexity. Besides previously reported softness and stretchability, desired merits for practical use include elasticity, solvent resistance, facile patternability and high charge carrier mobility. Here, we show a molecular design concept that simultaneously achieves all these targeted properties in both polymeric semiconductors and dielectrics, without compromising electrical performance. This is enabled by covalently-embedded in-situ rubber matrix (iRUM) formation through good mixing of iRUM precursors with polymer electronic materials, and finely-controlled composite film morphology built on azide crosslinking chemistry which leverages different reactivities with C–H and C=C bonds. The high covalent crosslinking density results in both superior elasticity and solvent resistance. When applied in stretchable transistors, the iRUM-semiconductor film retained its mobility after stretching to 100% strain, and exhibited record-high mobility retention of 1 cm 2 V −1 s −1 after 1000 stretching-releasing cycles at 50% strain. The cycling life was stably extended to 5000 cycles, five times longer than all reported semiconductors. Furthermore, we fabricated elastic transistors via consecutively photo-patterning of the dielectric and semiconducting layers, demonstrating the potential of solution-processed multilayer device manufacturing. The iRUM represents a molecule-level design approach towards robust skin-inspired electronics. 

Abstract Schizophrenia is a severe and complex psychiatric disorder with heterogeneous and dynamic multi-dimensional symptoms. Behavioral rhythms, such as sleep rhythm, are usually disrupted in people with schizophrenia. As such, behavioral rhythm sensing with smartphones and machine learning can help better understand and predict their symptoms. Our goal is to predict fine-grained symptom changes with interpretable models. We computed rhythm-based features from 61 participants with 6,132 days of data and used multi-task learning to predict their ecological momentary assessment scores for 10 different symptom items. By taking into account both the similarities and differences between different participants and symptoms, our multi-task learning models perform statistically significantly better than the models trained with single-task learning for predicting patients’ individual symptom trajectories, such as feeling depressed, social, and calm and hearing voices. We also found different subtypes for each of the symptoms by applying unsupervised clustering to the feature weights in the models. Taken together, compared to the features used in the previous studies, our rhythm features not only improved models’ prediction accuracy but also provided better interpretability for how patients’ behavioral rhythms and the rhythms of their environments influence their symptom conditions. This will enable both the patients and clinicians to monitor how these factors affect a patient’s condition and how to mitigate the influence of these factors. As such, we envision that our solution allows early detection and early intervention before a patient’s condition starts deteriorating without requiring extra effort from patients and clinicians. 

Abstract The majority of massive star-forming galaxies atz ∼ 2 have velocity gradients suggestive of rotation, in addition to large amounts of disordered motions. In this paper, we demonstrate that it is challenging to distinguish the regular rotation of a disk galaxy from the orbital motions of merging galaxies with seeing-limited data. However, the merger fractions atz ∼ 2 are likely too low for this to have a large effect on measurements of disk fractions. To determine how often mergers pass for disks, we look to galaxy formation simulations. We analyze ∼24,000 synthetic images and kinematic maps of 31 high-resolution simulations of isolated galaxies and mergers atz ∼ 2. We determine if the synthetic observations pass the criteria commonly used to identify disk galaxies and whether the results are consistent with their intrinsic dynamical states. Galaxies that are intrinsically mergers pass the disk criteria for anywhere from 0% to 100% of sightlines. The exact percentage depends strongly on the specific disk criteria adopted and weakly on the separation of the merging galaxies. Therefore, one cannot tell with certainty whether observations of an individual galaxy indicate a merger or a disk. To estimate the fraction of mergers passing as disks in current kinematics samples, we combine the probability that a merger will pass as a disk with theoretical merger fractions from a cosmological simulation. Taking the latter at face value, the observed disk fractions are overestimated by small amounts: at most by 5% at high stellar mass (10^10–11M_⊙) and 15% at low stellar mass (10^9–10M_⊙). 

Abstract In order to apply polymer semiconductors to stretchable electronics, they need to be easily deformed under strain without being damaged. A small number of conjugated polymers, typically with semicrystalline packing structures, have been reported to exhibit mechanical stretchability. Herein, a method is reported to modify polymer semiconductor packing‐structure using a molecular additive, dioctyl phthalate (DOP), which is found to act as a molecular spacer, to be inserted between the amorphous chain networks and disrupt the crystalline packing. As a result, large‐crystal growth is suppressed while short‐range aggregations of conjugated polymers are promoted, which leads to an improved mechanical stretchability without affecting charge‐carrier transport. Due to the reduced conjugated polymer intermolecular interactions, strain‐induced chain alignment and crystallization are observed. By adding DOP to a well‐known conjugated polymer, poly[2,5‐bis(4‐decyltetradecyl)pyrrolo[3,4‐c]pyrrole‐1,4‐(2H,5H)‐dione‐(E)‐1,2‐di(2,2′‐bithiophen‐5‐yl)ethene] (DPPTVT), stretchable transistors are obtained with anisotropic charge‐carrier mobilities under strain, and stable current output under strain up to 100%.