Search

IntroductionCleidocranial dysplasia (CCD) is an autosomal-dominant, heritable skeletal and dental disease, involving hypoplastic clavicles, defective ossification of the anterior fontanelle, dentin and enamel hypoplasia, and supernumerary teeth, which can seriously affect the oral and mental health of patients. Amyloid-like protein aggregation, which is established by lysozyme conjugated with polyethylene glycol (Lyso-PEG), forms a mineralized nanofilm layer on a healthy enamel surface. However, whether it can form a remineralization layer in dental tissues from CCD remains unclear.MethodsThis study evaluated deciduous teeth from healthy individuals and a patient with CCD. Because pulp and dentin are functionally closely related, stem cells from human exfoliated deciduous teeth (SHED) from CCD patients and healthy individuals were collected to compare their biological properties.ResultsThe results found that deciduous teeth from patients with CCD exhibited dentin hypoplasia. In addition, the proliferative ability and osteogenic potential of SHED from patients with CCD were lower than those of control individuals. Finally, Lyso-PEG was applied to dentin from the CCD and control groups, showing a similar remineralization-induced effect on the dentin surfaces of the two groups.ConclusionThese results extend our understanding of the dentin and SHED of patients with CCD, exhibiting good caries-preventive capacity and good biocompatibility of Lyso-PEG, thus providing a novel dental therapy for CCD and patients with tooth hypoplasia.

In statistics, samples are drawn from a population in a data- generating process (DGP). Standard errors measure the uncer- tainty in sample estimates of population parameters. In sci- ence, evidence is generated to test hypotheses in an evidence- generating process (EGP). We claim that EGP variation across researchers adds uncertainty: non-standard errors. To study them, we let 164 teams test six hypotheses on the same sam- ple. We find that non-standard errors are sizeable, on par with standard errors. Their size (i) co-varies only weakly with team merits, reproducibility, or peer rating, (ii) declines significantly after peer-feedback, and (iii) is underestimated by participants.

To foster the development of new models for collaborative AI-assisted report generation, we introduce MegaWika, consisting of 13 million Wikipedia articles in 50 diverse languages, along with their 71 million referenced source materials. We process this dataset for a myriad of applications, going beyond the initial Wikipedia citation extraction and web scraping of content, including translating non-English articles for cross-lingual applications and providing FrameNet parses for automated semantic analysis. MegaWika is the largest resource for sentence-level report generation and the only report generation dataset that is multilingual. We manually analyze the quality of this resource through a semantically stratified sample. Finally, we provide baseline results and trained models for crucial steps in automated report generation: cross-lingual question answering and citation retrieval., Comment: Submitted to ACL, 2023

Supervised learning is often affected by a covariate shift in which the marginal distributions of instances (covariates $x$) of training and testing samples $\mathrm{p}_\text{tr}(x)$ and $\mathrm{p}_\text{te}(x)$ are different but the label conditionals coincide. Existing approaches address such covariate shift by either using the ratio $\mathrm{p}_\text{te}(x)/\mathrm{p}_\text{tr}(x)$ to weight training samples (reweighted methods) or using the ratio $\mathrm{p}_\text{tr}(x)/\mathrm{p}_\text{te}(x)$ to weight testing samples (robust methods). However, the performance of such approaches can be poor under support mismatch or when the above ratios take large values. We propose a minimax risk classification (MRC) approach for covariate shift adaptation that avoids such limitations by weighting both training and testing samples. In addition, we develop effective techniques that obtain both sets of weights and generalize the conventional kernel mean matching method. We provide novel generalization bounds for our method that show a significant increase in the effective sample size compared with reweighted methods. The proposed method also achieves enhanced classification performance in both synthetic and empirical experiments.

It can be difficult to identify trends and perform quality control in large, high-dimensional fMRI or omics datasets. To remedy this, we develop ImageNomer, a data visualization and analysis tool that allows inspection of both subject-level and cohort-level features. The tool allows visualization of phenotype correlation with functional connectivity (FC), partial connectivity (PC), dictionary components (PCA and our own method), and genomic data (single-nucleotide polymorphisms, SNPs). In addition, it allows visualization of weights from arbitrary ML models. ImageNomer is built with a Python backend and a Vue frontend. We validate ImageNomer using the Philadelphia Neurodevelopmental Cohort (PNC) dataset, which contains multitask fMRI and SNP data of healthy adolescents. Using correlation, greedy selection, or model weights, we find that a set of 10 FC features can explain 15% of variation in age, compared to 35% for the full 34,716 feature model. The four most significant FCs are either between bilateral default mode network (DMN) regions or spatially proximal subcortical areas. Additionally, we show that whereas both FC (fMRI) and SNPs (genomic) features can account for 10-15% of intelligence variation, this predictive ability disappears when controlling for race. We find that FC features can be used to predict race with 85% accuracy, compared to 78% accuracy for sex prediction. Using ImageNomer, this work casts doubt on the possibility of finding unbiased intelligence-related features in fMRI and SNPs of healthy adolescents., Comment: 11 pages

Prevalent deterministic deep-learning models suffer from significant over-confidence under distribution shifts. Probabilistic approaches can reduce this problem but struggle with computational efficiency. In this paper, we propose Density-Softmax, a fast and lightweight deterministic method to improve calibrated uncertainty estimation via a combination of density function with the softmax layer. By using the latent representation's likelihood value, our approach produces more uncertain predictions when test samples are distant from the training samples. Theoretically, we show that Density-Softmax can produce high-quality uncertainty estimation with neural networks, as it is the solution of minimax uncertainty risk and is distance-aware, thus reducing the over-confidence of the standard softmax. Empirically, our method enjoys similar computational efficiency as a single forward pass deterministic with standard softmax on the shifted toy, vision, and language datasets across modern deep-learning architectures. Notably, Density-Softmax uses 4 times fewer parameters than Deep Ensembles and 6 times lower latency than Rank-1 Bayesian Neural Network, while obtaining competitive predictive performance and lower calibration errors under distribution shifts., Comment: 37 pages, 6 tables, 22 figures

In the flow distribution of oil and gas gathering and transportation system, the flow pattern, inertia force, gravity, and other factors will lead to uneven flow distribution of each outlet pipe, which will seriously affect the economy and safety of the gathering and transportation system. Therefore, it is of great significance to analyze the characteristics of gas–liquid two-phase flow in horizontal pipelines. For this reason, an experimental device for maldistribution control of gas–liquid two-phase flow is developed in this article. In order to solve the problem of flow maldistribution in gas–liquid two-phase flow pipe, the test experiments of maldistribution control under the conditions of three flow patterns (stratified flow, slug flow, and annular flow) are carried out. Through experiments, we revealed the control law of flow pattern to liquid maldistribution and gas maldistribution with or without flute pipes and formed the maldistribution control scheme of gas–liquid two-phase flow in the horizontal pipeline. It is found that it is more beneficial to the control of average liquid maldistribution degree under the condition of slug flow pattern in horizontal manifold without flute pipe, and it is more beneficial to the control of average gas maldistribution degree under the condition of annular flow pattern without flute pipe. Without the function of the flute pipe and in the annular flow pattern, it is beneficial to the control of average maldistribution degree of gas–liquid two-phase flow. The findings of this research can be used as a reference for the field maldistribution control of oil and gas gathering and transportation systems.

We study the problem of invariant learning when the environment labels are unknown. We focus on the invariant representation notion when the Bayes optimal conditional label distribution is the same across different environments. Previous work conducts Environment Inference (EI) by maximizing the penalty term from Invariant Risk Minimization (IRM) framework. The EI step uses a reference model which focuses on spurious correlations to efficiently reach a good environment partition. However, it is not clear how to find such a reference model. In this work, we propose to repeat the EI process and retrain an ERM model on the \textit{majority} environment inferred by the previous EI step. Under mild assumptions, we find that this iterative process helps learn a representation capturing the spurious correlation better than the single step. This results in better Environment Inference and better Invariant Learning. We show that this method outperforms baselines on both synthetic and real-world datasets., Comment: ICML workshop on Spurious Correlations, Invariance and Stability (2022)

We propose \textbf{JAWS}, a series of wrapper methods for distribution-free uncertainty quantification tasks under covariate shift, centered on the core method \textbf{JAW}, the \textbf{JA}ckknife+ \textbf{W}eighted with data-dependent likelihood-ratio weights. JAWS also includes computationally efficient \textbf{A}pproximations of JAW using higher-order influence functions: \textbf{JAWA}. Theoretically, we show that JAW relaxes the jackknife+'s assumption of data exchangeability to achieve the same finite-sample coverage guarantee even under covariate shift. JAWA further approaches the JAW guarantee in the limit of the sample size or the influence function order under common regularity assumptions. Moreover, we propose a general approach to repurposing predictive interval-generating methods and their guarantees to the reverse task: estimating the probability that a prediction is erroneous, based on user-specified error criteria such as a safe or acceptable tolerance threshold around the true label. We then propose \textbf{JAW-E} and \textbf{JAWA-E} as the repurposed proposed methods for this \textbf{E}rror assessment task. Practically, JAWS outperform state-of-the-art predictive inference baselines in a variety of biased real world data sets for interval-generation and error-assessment predictive uncertainty auditing tasks., Comment: Thirty-sixth Conference on Neural Information Processing Systems

Contemporary vision benchmarks predominantly consider tasks on which humans can achieve near-perfect performance. However, humans are frequently presented with visual data that they cannot classify with 100% certainty, and models trained on standard vision benchmarks achieve low performance when evaluated on this data. To address this issue, we introduce a procedure for creating datasets of ambiguous images and use it to produce SQUID-E ("Squidy"), a collection of noisy images extracted from videos. All images are annotated with ground truth values and a test set is annotated with human uncertainty judgments. We use this dataset to characterize human uncertainty in vision tasks and evaluate existing visual event classification models. Experimental results suggest that existing vision models are not sufficiently equipped to provide meaningful outputs for ambiguous images and that datasets of this nature can be used to assess and improve such models through model training and direct evaluation of model calibration. These findings motivate large-scale ambiguous dataset creation and further research focusing on noisy visual data., Comment: 10 pages, NeurIPS 2022 Datasets and Benchmarks Track

To meet the increasing wastewater treatment demand while minimizing the land footprint of the treatment systems and plants, more efficient and compact processes are needed. The circulating fluidized bed bioreactor (CFBBR) has been proven to achieve high levels of biological nutrient removal. Past studies at the lab and pilot scale achieved 94% COD removal and 80% nitrogen removal at HRT's of 2-4 h. A collaborative project between Western University and the Guangzhou Institute of Energy Conversion (GIEC), in Guangzhou, China, further explored the treatment of municipal wastewater with the CFBBR. A pilot CFBBR, with aerobic and anoxic columns for nitrification and denitrification, was constructed at the GIEC for in-situ treatment of septic tank effluent from a residential building. Due to high concentrations of ammonia (NH4-N), the wastewater had a COD/N ratio of 2-3. Thus, operating at a longer HRT and supplementing COD, in the form of glucose, was necessary to achieve a high nitrogen removal efficiency. The system was run both with and without supplemental COD at HRT's between 16 and 21 h, treating approximately 1000-1270 L/d. Overall, a COD removal efficiency of at least 92%, ammonia removal of 97%, and nitrogen removal of 82% was achieved. The CFBBR system achieved an effluent with BOD and NH4-N concentrations both below 5 mg/L, a NO3-N concentration below 15 mg/L, and a total nitrogen concentration below 25 mg/L. The compact design of this pilot-CFBBR, coupled with its high BNR performance make it an excellent option for decentralized treatment of urban wastewaters.

Wind environment as an essential aspect of urban micro-climate is usually studied as an important factor affecting human wind comfort and thermal comfort in public open space. Less studied is how wind environments influence schoolyard serving for children and teenagers who are more sensitive to air temperature and air pollution. This paper presents a study using XFlow for wind speed evaluation based on CFD (Computational Fluid Dynamics) simulation, and Ecotect Analysis for thermal comfort evaluation. Choosing Nanjing Mochou Lake Primary School and Nanjing NO.13 Middle School Suojin Campus as case studies, it offers an evaluation of wind environments in these two schoolyards and surrounding open spaces. Simulation shows that improvements should be made to build better wind environments around both schools. Scenarios are developed subsequently by adjusting buildings in study areas at the scale of blocks and school buildings, proving the efficiency of improvements for a more comfortable wind environments. This study focuses on schoolyard wind environments from the standpoint of children and teenagers, suggesting that research on urban micro-climate can be expanded in multiple directions so that as many social groups in different ages as possible benefits, i.e. have healthier urban life.

We address the problem of active learning under label shift: when the class proportions of source and target domains differ. We introduce a "medial distribution" to incorporate a tradeoff between importance weighting and class-balanced sampling and propose their combined usage in active learning. Our method is known as Mediated Active Learning under Label Shift (MALLS). It balances the bias from class-balanced sampling and the variance from importance weighting. We prove sample complexity and generalization guarantees for MALLS which show active learning reduces asymptotic sample complexity even under arbitrary label shift. We empirically demonstrate MALLS scales to high-dimensional datasets and can reduce the sample complexity of active learning by 60% in deep active learning tasks., 18 pages, 9 figures, to appear at the 2021 International Conference on Artificial Intelligence and Statistics (AIStats)

Supplemental Material, sj-pdf-1-cpc-10.1177_10556656211018950 for Mechanical Tension-Stress in Alveolar Cleft Repaired With Autogenous Bone in Canine Models by Anqi Liu and Jialiang Huang in The Cleft Palate-Craniofacial Journal

The financial market is continuously shaped by the interplay between regulations and the actions from market participants. This dissertation investigates how market design impacts market landscape, by providing systematic incentives for certain player behaviours. Three papers examining three different policy changes are included. In the first study, I investigate one mechanism that contributed to the current fragmented North American equity marketplace. I use the introduction of the Order Protection Rule in 2010 as a natural experiment to examine the mechanism behind such proliferation of small venues. We find that 'queue jumping' mechanism can act as a starter to the liquidity on small or empty trading venues. In the second study, I examine an attempt by a stock exchange to bring liquidity from off-market, such as over-the-counter, to on-market continuous trading. The change is through removing the fixed fee per trade. Such change encourages order slicing and optimal execution through algorithms, thus showing significant impact on institutional trading. In the third study, I study an innovative method to address one of the most prominent issues in the modern equity market, liquidity for small-cap stocks. The solution proposed is to let liquid stocks cross-subsidise the illiquid ones in the Designated Market Makers program. Through investigating market makers' profit structure and incentive, I find two mechanisms through which the program encourages them to quote tighter spreads and higher depths. This dissertation contributed to the discussions around several of the most important topics in the modern equity market, including fragmentation, institutional optimal execution strategies and liquidity for small-cap stocks. By researching the channels driving these changes, these papers provide policy implications that reach far beyond the three policy changes that are directly examined.

We propose a framework for learning calibrated uncertainties under domain shifts. We consider the case where the source (training) distribution differs from the target (test) distribution. We detect such domain shifts through the use of a binary domain classifier and integrate it with the task network and train them jointly end-to-end. The binary domain classifier yields a density ratio that reflects the closeness of a target (test) sample to the source (training) distribution. We employ it to adjust the uncertainty of prediction in the task network. This idea of using the density ratio is based on the distributionally robust learning (DRL) framework, which accounts for the domain shift through adversarial risk minimization. We demonstrate that our method generates calibrated uncertainties that benefit many downstream tasks, such as unsupervised domain adaptation (UDA) and semi-supervised learning (SSL). In these tasks, methods like self-training and FixMatch use uncertainties to select confident pseudo-labels for re-training. Our experiments show that the introduction of DRL leads to significant improvements in cross-domain performance. We also demonstrate that the estimated density ratios show agreement with the human selection frequencies, suggesting a positive correlation with a proxy of human perceived uncertainties., Preprint

In this study, we use entropy-based measures to identify different types of trading behaviors. We detect the return-driven trading using the conditional block entropy that dynamically reflects the &ldquo, self-causality&rdquo, of market return flows. Then we use the transfer entropy to identify the news-driven trading activity that is revealed by the information flows from news sentiment to market returns. We argue that when certain trading behavior becomes dominant or jointly dominant, the market will form a specific regime, namely return-, news- or mixed regime. Based on 11 years of news and market data, we find that the evolution of financial market regimes in terms of adaptive trading activities over the 2008 liquidity and euro-zone debt crises can be explicitly explained by the information flows. The proposed method can be expanded to make &ldquo, causal&rdquo, inferences on other types of economic phenomena.

Conversion of lignocellulose to sugars suitable for microbial fermentation is an outstanding obstacle in developing biorefinery. Both hemicellulose and cellulose fractions are polymers of sugars and thereby primary candidates for fermentable sugars production. In this study, the flexibility of an integrated biomass conversion process was offered. The hemicellulose of corncobs was utilized to release fermentable sugars by sulfuric acid hydrolysis first. The remaining solid residue from acid hydrolysis, containing a lot of cellulose, was further used to produce levoglucosan by fast pyrolysis. This process appeared to present several advantages: (i) Almost all of hemicellulose (99.7%) was hydrolyzed, and the yield of xylan was achieved 86.1%. (ii) The alkali and alkaline earth metals, which had negative catalytic influence on levoglucosan formation, were nearly and completely (93.7%) removed by acid pretreatment. (iii) A preferential degradation of hemicellulose and amorphous cellulose during acid hydrolysis...

In this work, a novel catalyst of Fe3O4@SiO2@Yb2O3 was prepared and the degradation of thymol in reverse osmosis concentrate using ozonation was explored. The operational parameters, such as ozone dosage (8–48 mg·min− 1), initial thymol concentration (20–100 mg·L− 1), initial pH value (3–11), and catalyst Fe3O4@SiO2@Yb2O3 dosage (0.2–1.0 g), were studied focusing on the thymol degradation and COD removal. The results indicated that the increase in ozone dosage, initial pH value, and Fe3O4@SiO2@Yb2O3 dosage accelerated the thymol degradation and COD removal, while the increase in initial thymol concentration hampered the effect of ozonation. A pathway of thymol degradation by catalytic ozonation was proposed based on the intermediates detected by gas chromatography-mass spectrometer and ion chromatography. This paper can provide basic data and technical alternative for pollutant removal from reverse osmosis concentrate by ozonation.

Online harassment is a significant social problem. Prevention of online harassment requires rapid detection of harassing, offensive, and negative social media posts. In this paper, we propose the use of word embedding models to identify offensive and harassing social media messages in two aspects: detecting fast-changing topics for more effective data collection and representing word semantics in different domains. We demonstrate with preliminary results that using the GloVe (Global Vectors for Word Representation) model facilitates the discovery of new and relevant keywords to use for data collection and trolling detection. Our paper concludes with a discussion of a research agenda to further develop and test word embedding models for identification of social media harassment and trolling., AI for Social Good workshop at NeurIPS (2019)

The studies presented in this dissertation aim to better understand the phonological status and phonetic realization of the rhotic suffix in modern Mandarin. The final result is a descriptive analysis of the phonological behaviour of the rhotic suffix, based on production studies. The thesis traces the historical evolution of the suffix, marked by its semantic bleaching, and focuses on the production studies, which examine the realization of the rhotic suffix in different contexts, according to the rhyme of the syllable to which it is attached. The results of these studies complement those of existing studies and allow to test analyses proposed in the literature on Mandarin phonology. On the basis of the results of the acoustic and articulatory studies presented here, a new phonological analysis of the rhotic suffix is proposed, in the framework of articulatory phonology (Browman & Goldstein, 1992), a model that takes the articulatory gesture as primitive. This analysis allows a more unified treatment of all the realizations of the rhotic suffix, in all the contexts where it appears., Les études présentées dans cette thèse visent à mieux comprendre le statut phonologique et la réalisation phonétique du suffixe rhotique en mandarin moderne. Le résultat final est une analyse descriptive du comportement phonologique du suffixe rhotique, basée sur des études de production. La thèse retrace l’évolution historique du suffixe marquée par sa désémantisation et se concentre sur l’étude de production, qui examine la réalisation du suffixe rhotique dans différents contextes, en fonction de la rime de la syllabe à laquelle il s’attache. Les résultats de ces études viennent compléter ceux des études déjà existantes et permettent de tester des analyses proposées dans la littérature sur la phonologie du mandarin. Sur la base des résultats de l’étude acoustique et de l’étude articulatoire présentées ici, une nouvelle analyse phonologique du suffixe rhotique est proposée dans le cadre de la phonologie articulatoire (Browman & Goldstein, 1992), modèle qui prend le geste articulatoire comme primitive. Cette analyse permet un traitement plus unifié de toutes les réalisations du suffixe rhotique, dans tous les contextes où il apparait.

We propose a robust regression approach to off-policy evaluation (OPE) for contextual bandits. We frame OPE as a covariate-shift problem and leverage modern robust regression tools. Ours is a general approach that can be used to augment any existing OPE method that utilizes the direct method. When augmenting doubly robust methods, we call the resulting method Triply Robust. We prove upper bounds on the resulting bias and variance, as well as derive novel minimax bounds based on robust minimax analysis for covariate shift. Our robust regression method is compatible with deep learning, and is thus applicable to complex OPE settings that require powerful function approximators. Finally, we demonstrate superior empirical performance across the standard OPE benchmarks, especially in the case where the logging policy is unknown and must be estimated from data., Comment: Preliminary Work

We model human decision-making behaviors in a risk-taking task using inverse reinforcement learning (IRL) for the purposes of understanding real human decision making under risk. To the best of our knowledge, this is the first work applying IRL to reveal the implicit reward function in human risk-taking decision making and to interpret risk-prone and risk-averse decision-making policies. We hypothesize that the state history (e.g. rewards and decisions in previous trials) are related to the human reward function, which leads to risk-averse and risk-prone decisions. We design features that reflect these factors in the reward function of IRL and learn the corresponding weight that is interpretable as the importance of features. The results confirm the sub-optimal risk-related decisions of human-driven by the personalized reward function. In particular, the risk-prone person tends to decide based on the current pump number, while the risk-averse person relies on burst information from the previous trial and the average end status. Our results demonstrate that IRL is an effective tool to model human decision-making behavior, as well as to help interpret the human psychological process in risk decision-making., Comment: Real-world Sequential Decision Making Workshop at ICML 2019

We study the problem of safe learning and exploration in sequential control problems. The goal is to safely collect data samples from operating in an environment, in order to learn to achieve a challenging control goal (e.g., an agile maneuver close to a boundary). A central challenge in this setting is how to quantify uncertainty in order to choose provably-safe actions that allow us to collect informative data and reduce uncertainty, thereby achieving both improved controller safety and optimality. To address this challenge, we present a deep robust regression model that is trained to directly predict the uncertainty bounds for safe exploration. We derive generalization bounds for learning, and connect them with safety and stability bounds in control. We demonstrate empirically that our robust regression approach can outperform the conventional Gaussian process (GP) based safe exploration in settings where it is difficult to specify a good GP prior., Comment: 2nd Annual Conference on Learning for Dynamics and Control

Prior researches have found that emotions and subsequent behaviours are closely bonded. Following that, in the context of global warming, this study examines how approach-oriented emotions (e.g. happiness) and avoidant emotions (e.g. sadness) influence the preference of energy, choice of diet and willingness to recycle. Thus, we have two conditions (i.e. happy and sad) in the study and each condition is asked to complete a survey regarding the pro-environmental behaviors. The result reveals that happiness makes people more likely to recycle and sadness has the opposite effect. However, no results are found in the willingness to use green energy and plant-based diet. This result contributes to the previous finding and give a dimensional perspective on how to analysis the effect of emotions on global warming actions. Limitations and implication are also discussed. Disclaimer: “UBC SEEDS provides students with the opportunity to share the findings of their studies, as well as their opinions, conclusions and recommendations with the UBC community. The reader should bear in mind that this is a student project/report and is not an official document of UBC. Furthermore readers should bear in mind that these reports may not reflect the current status of activities at UBC. We urge you to contact the research persons mentioned in a report or the SEEDS Coordinator about the current status of the subject matter of a project/report.”

We propose a robust adversarial prediction framework for general multiclass classification. Our method seeks predictive distributions that robustly optimize non-convex and non-continuous multiclass loss metrics against the worst-case conditional label distributions (the adversarial distributions) that (approximately) match the statistics of the training data. Although the optimized loss metrics are non-convex and non-continuous, the dual formulation of the framework is a convex optimization problem that can be recast as a risk minimization model with a prescribed convex surrogate loss we call the adversarial surrogate loss. We show that the adversarial surrogate losses fill an existing gap in surrogate loss construction for general multiclass classification problems, by simultaneously aligning better with the original multiclass loss, guaranteeing Fisher consistency, enabling a way to incorporate rich feature spaces via the kernel trick, and providing competitive performance in practice., 49 pages, 10 figures

Raw syngas conditioning using molten salts was carried out in a fixed-bed reactor. The effects of the reaction conditions, including temperature, gas velocity, bubble diameter, molten salt static liquid heights, and inlet gas composition, on the composition of the syngas product and the properties of the spent molten salts were investigated. The molten salt absorbed CO2 in all of the experiments (at temperatures from 350 &deg, C to 500 &deg, C) and decreased its concentration down to 1%. The H2/CO increased from 0.94 to a maximum of 11.0, which may meet most of the synthetic process&rsquo, requirements. The temperature, gas velocity, and molten salt static liquid heights had significant effects on the H2 and CO concentrations in the gas product. Molten salt raw syngas conditioning can be a suitable follow-up procedure for gasification and pyrolysis.

Under covariate shift, training (source) data and testing (target) data differ in input space distribution, but share the same conditional label distribution. This poses a challenging machine learning task. Robust Bias-Aware (RBA) prediction provides the conditional label distribution that is robust to the worstcase logarithmic loss for the target distribution while matching feature expectation constraints from the source distribution. However, employing RBA with insufficient feature constraints may result in high certainty predictions for much of the source data, while leaving too much uncertainty for target data predictions. To overcome this issue, we extend the representer theorem to the RBA setting, enabling minimization of regularized expected target risk by a reweighted kernel expectation under the source distribution. By applying kernel methods, we establish consistency guarantees and demonstrate better performance of the RBA classifier than competing methods on synthetically biased UCI datasets as well as datasets that have natural covariate shift.

Covariate shift relaxes the widely-employed independent and identically distributed (IID) assumption by allowing different training and testing input distributions. Unfortunately, common methods for addressing covariate shift by trying to remove the bias between training and testing distributions using importance weighting often provide poor performance guarantees in theory and unreliable predictions with high variance in practice. Recently developed methods that construct a predictor that is inherently robust to the difficulties of learning under covariate shift are restricted to minimizing logloss and can be too conservative when faced with high-dimensional learning tasks. We address these limitations in two ways: by robustly minimizing various loss functions, including non-convex ones, under the testing distribution; and by separately shaping the influence of covariate shift according to different feature-based views of the relationship between input variables and example labels. These generalizations make robust covariate shift prediction applicable to more task scenarios. We demonstrate the benefits on classification under covariate shift tasks.

MicroRNA-9-1(miR-9-1) plays an important role in the mechanism that regulates the lineage fate of differentiating hematopoietic cells. Recent studies have shown that miR-9-1 is downregulated in t (8; 21) AML. However, the pathogenic mechanisms underlying miR-9-1 downregulation and the RUNX1-RUNX1T1 fusion protein, generated from the translocation of t (8; 21) in AML, remain unclear. RUNX1-RUNX1T1 can induce leukemogenesis through resides in and functions as a stable RUNX1-RUNX1T1-containing transcription factor complex. In this study, we demonstrate that miR-9-1 expression increases significantly after the treatment of RUNX1-RUNX1T1 (+) AML cell lines with decitabine (a DNMT inhibitor) and trichostatin A (an HDAC inhibitor). In addition, we show that RUNX1-RUNX1T1 triggers the heterochromatic silencing of miR-9-1 by binding to RUNX1-binding sites in the promoter region of miR-9-1 and recruiting chromatin-remodeling enzymes, DNMTs, and HDACs, contributing to hypermethylation of miR-9-1 in t (8; 21) AML. Furthermore, because RUNX1, RUNX1T1, and RUNX1-RUNX1T1 are all regulated by miR-9-1, the silencing of miR-9-1 enhances the oncogenic activity of these genes. Besides, overexpression of miR-9-1 induces differentiation and inhibits proliferation in t (8; 21) AML cell lines. In conclusion, our results indicate a feedback circuitry involving miR-9-1 and RUNX1-RUNX1T1, contributing to leukemogenesis in RUNX1-RUNX1T1 (+) AML cell lines.