PyTorchKR Papers

최신 논문

총 739편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2106.00719

Stochastic Collapsed Variational Inference for Structured Gaussian Process Regression Network

This paper presents an efficient variational inference framework for deriving a family of structured gaussian process regression network (SGPRN) models. The key idea is to incorporate auxiliary inducing variables in latent functions and jointly treats both the distributions of the inducing variables and hyper-parameters as variational parameters. Then we propose structured variable distributions and marginalize latent variables, which enables the decomposability of a tractable variational lower bound and leads to stochastic optimization. Our inference approach is able to model data in which outputs do not share a common input set with a computational complexity independent of the size of the inputs and outputs and thus easily handle datasets with missing values. We illustrate the performance of our method on synthetic data and real datasets and show that our model generally provides better imputation results on missing data than the state-of-the-art. We also provide a visualization approach for time-varying correlation across outputs in electrocoticography data and those estimates provide insight to understand the neural population dynamics.

arXiv:2106.00614

Pattern Discovery in Time Series with Byte Pair Encoding

The growing popularity of wearable sensors has generated large quantities of temporal physiological and activity data. Ability to analyze this data offers new opportunities for real-time health monitoring and forecasting. However, temporal physiological data presents many analytic challenges: the data is noisy, contains many missing values, and each series has a different length. Most methods proposed for time series analysis and classification do not handle datasets with these characteristics nor do they offer interpretability and explainability, a critical requirement in the health domain. We propose an unsupervised method for learning representations of time series based on common patterns identified within them. The patterns are, interpretable, variable in length, and extracted using Byte Pair Encoding compression technique. In this way the method can capture both long-term and short-term dependencies present in the data. We show that this method applies to both univariate and multivariate time series and beats state-of-the-art approaches on a real world dataset collected from wearable sensors.

arXiv:2106.00002

Analysis and classification of main risk factors causing stroke in Shanxi Province

In China, stroke is the first leading cause of death in recent years. It is a major cause of long-term physical and cognitive impairment, which bring great pressure on the National Public Health System. Evaluation of the risk of getting stroke is important for the prevention and treatment of stroke in China. A data set with 2000 hospitalized stroke patients in 2018 and 27583 residents during the year 2017 to 2020 is analyzed in this study. Due to data incompleteness, inconsistency, and non-structured formats, missing values in the raw data are filled with -1 as an abnormal class. With the cleaned features, three models on risk levels of getting stroke are built by using machine learning methods. The importance of "8+2" factors from China National Stroke Prevention Project (CSPP) is evaluated via decision tree and random forest models. Except for "8+2" factors the importance of features and SHAP1 values for lifestyle information, demographic information, and medical measurement are evaluated and ranked via a random forest model. Furthermore, a logistic regression model is applied to evaluate the probability of getting stroke for different risk levels. Based on the census data in both communities and hospitals from Shanxi Province, we investigate different risk factors of getting stroke and their ranking with interpretable machine learning models. The results show that Hypertension (Systolic blood pressure, Diastolic blood pressure), Physical Inactivity (Lack of sports), and Overweight (BMI) are ranked as the top three high-risk factors of getting stroke in Shanxi province. The probability of getting stroke for a person can also be predicted via our machine learning model.

arXiv:2106.00311공식 코드 · Papers with Code

What's a good imputation to predict with missing values?

How to learn a good predictor on data with missing values? Most efforts focus on first imputing as well as possible and second learning on the completed data to predict the outcome. Yet, this widespread practice has no theoretical grounding. Here we show that for almost all imputation functions, an impute-then-regress procedure with a powerful learner is Bayes optimal. This result holds for all missing-values mechanisms, in contrast with the classic statistical results that require missing-at-random settings to use imputation in probabilistic modeling. Moreover, it implies that perfect conditional imputation is not needed for good prediction asymptotically. In fact, we show that on perfectly imputed data the best regression function will generally be discontinuous, which makes it hard to learn. Crafting instead the imputation so as to leave the regression function unchanged simply shifts the problem to learning discontinuous imputations. Rather, we suggest that it is easier to learn imputation and regression jointly. We propose such a procedure, adapting NeuMiss, a neural network capturing the conditional links across observed and unobserved variables whatever the missing-value pattern. Experiments confirm that joint imputation and regression through NeuMiss is better than various two step procedures in our experiments with finite number of samples.

arXiv:2105.08864

Detection of Multidecadal Changes in Vegetation Dynamics and Association with Intra-annual Climate Variability in the Columbia River Basin

Leaf Area index is widely used metric for the assessment of vegetation dynamics and can be used to assess the impact of regional/local climate conditions. The underlying continuity of high resolution spatio-temporal phenological processes in the presence of extensive missing values poses a number of challenges in the detection of changes at a local and regional level. The feasibility of functional data analysis methods were evaluated to improve the exploration of such data. In this paper, an investigation of multidecadal variation of leaf area index (LAI) is conducted in the Columbia Watershed, as detected by NOAA AVHRR satellite imaging, and its inter- and intra-annual correlation with maximum temperature and precipitation using the ERA-Interim Reanalysis from 1996 to 2017. A functional cluster analysis model was implemented to identify regions in the Columbia Watershed that exhibit similar long-term greening trends. Across these several regions, the primary source of annual LAI variation is a trend toward seasonally earlier and higher recordings of regional average maximum LAI. Further exploratory analysis reveals that although strongly correlated to LAI, maximum temperature and precipitation do not exhibit clear longitudinal trends.

arXiv:2105.06811공식 코드 · Papers with Code

Quantified Sleep: Machine learning techniques for observational n-of-1 studies

This paper applies statistical learning techniques to an observational Quantified-Self (QS) study to build a descriptive model of sleep quality. A total of 472 days of my sleep data was collected with an Oura ring and combined with lifestyle, environmental, and psychological data. Such n-of-1 QS projects pose a number of challenges: heterogeneous data sources; missing values; high dimensionality; dynamic feedback loops; human biases. This paper directly addresses these challenges with an end-to-end QS pipeline that produces robust descriptive models. Sleep quality is one of the most difficult modelling targets in QS research, due to high noise and a large number of weakly-contributing factors. Sleep quality was selected so that approaches from this paper would generalise to most other n-of-1 QS projects. Techniques are presented for combining and engineering features for the different classes of data types, sample frequencies, and schema - including event logs, weather, and geo-spatial data. Statistical analyses for outliers, normality, (auto)correlation, stationarity, and missing data are detailed, along with a proposed method for hierarchical clustering to identify correlated groups of features. The missing data was overcome using a combination of knowledge-based and statistical techniques, including several multivariate imputation algorithms. "Markov unfolding" is presented for collapsing the time series into a collection of independent observations, whilst incorporating historical information. The final model was interpreted in two ways: by inspecting the internal $\beta$-parameters, and using the SHAP framework. These two interpretation techniques were combined to produce a list of the 16 most-predictive features, demonstrating that an observational study can greatly narrow down the number of features that need to be considered when designing interventional QS studies.

arXiv:2105.06643공식 코드 · Papers with Code

Monash Time Series Forecasting Archive

Many businesses and industries nowadays rely on large quantities of time series data making time series forecasting an important research area. Global forecasting models that are trained across sets of time series have shown a huge potential in providing accurate forecasts compared with the traditional univariate forecasting models that work on isolated series. However, there are currently no comprehensive time series archives for forecasting that contain datasets of time series from similar sources available for the research community to evaluate the performance of new global forecasting algorithms over a wide variety of datasets. In this paper, we present such a comprehensive time series forecasting archive containing 20 publicly available time series datasets from varied domains, with different characteristics in terms of frequency, series lengths, and inclusion of missing values. We also characterise the datasets, and identify similarities and differences among them, by conducting a feature analysis. Furthermore, we present the performance of a set of standard baseline forecasting methods over all datasets across eight error metrics, for the benefit of researchers using the archive to benchmark their forecasting algorithms.

arXiv:2105.01733

Calibration of prediction rules for life-time outcomes using prognostic Cox regression survival models and multiple imputations to account for missing predictor data with cross-validatory assessment

In this paper, we expand the methodology presented in Mertens et. al (2020, Biometrical Journal) to the study of life-time (survival) outcome which is subject to censoring and when imputation is used to account for missing values. We consider the problem where missing values can occur in both the calibration data as well as newly - to-be-predicted - observations (validation). We focus on the Cox model. Methods are described to combine imputation with predictive calibration in survival modeling subject to censoring. Application to cross-validation is discussed. We demonstrate how conclusions broadly confirm the first paper which restricted to the study of binary outcomes only. Specifically prediction-averaging appears to have superior statistical properties, especially smaller predictive variation, as opposed to a direct application of Rubin's rules. Distinct methods for dealing with the baseline hazards are discussed when using Rubin's rules-based approaches.

arXiv:2104.14936공식 코드 · Papers with Code

Low-Rank Autoregressive Tensor Completion for Spatiotemporal Traffic Data Imputation

Spatiotemporal traffic time series (e.g., traffic volume/speed) collected from sensing systems are often incomplete with considerable corruption and large amounts of missing values, preventing users from harnessing the full power of the data. Missing data imputation has been a long-standing research topic and critical application for real-world intelligent transportation systems. A widely applied imputation method is low-rank matrix/tensor completion; however, the low-rank assumption only preserves the global structure while ignores the strong local consistency in spatiotemporal data. In this paper, we propose a low-rank autoregressive tensor completion (LATC) framework by introducing \textit{temporal variation} as a new regularization term into the completion of a third-order (sensor $\times$ time of day $\times$ day) tensor. The third-order tensor structure allows us to better capture the global consistency of traffic data, such as the inherent seasonality and day-to-day similarity. To achieve local consistency, we design the temporal variation by imposing an AR($p$) model for each time series with coefficients as learnable parameters. Different from previous spatial and temporal regularization schemes, the minimization of temporal variation can better characterize temporal generative mechanisms beyond local smoothness, allowing us to deal with more challenging scenarios such "blackout" missing. To solve the optimization problem in LATC, we introduce an alternating minimization scheme that estimates the low-rank tensor and autoregressive coefficients iteratively. We conduct extensive numerical experiments on several real-world traffic data sets, and our results demonstrate the effectiveness of LATC in diverse missing scenarios.

arXiv:2104.14914

BERT Meets Relational DB: Contextual Representations of Relational Databases

In this paper, we address the problem of learning low dimension representation of entities on relational databases consisting of multiple tables. Embeddings help to capture semantics encoded in the database and can be used in a variety of settings like auto-completion of tables, fully-neural query processing of relational joins queries, seamlessly handling missing values, and more. Current work is restricted to working with just single table, or using pretrained embeddings over an external corpus making them unsuitable for use in real-world databases. In this work, we look into ways of using these attention-based model to learn embeddings for entities in the relational database. We are inspired by BERT style pretraining methods and are interested in observing how they can be extended for representation learning on structured databases. We evaluate our approach of the autocompletion of relational databases and achieve improvement over standard baselines.

arXiv:2104.14278

ReLearn: A Robust Machine Learning Framework in Presence of Missing Data for Multimodal Stress Detection from Physiological Signals

Continuous and multimodal stress detection has been performed recently through wearable devices and machine learning algorithms. However, a well-known and important challenge of working on physiological signals recorded by conventional monitoring devices is missing data due to sensors insufficient contact and interference by other equipment. This challenge becomes more problematic when the user/patient is mentally or physically active or stressed because of more frequent conscious or subconscious movements. In this paper, we propose ReLearn, a robust machine learning framework for stress detection from biomarkers extracted from multimodal physiological signals. ReLearn effectively copes with missing data and outliers both at training and inference phases. ReLearn, composed of machine learning models for feature selection, outlier detection, data imputation, and classification, allows us to classify all samples, including those with missing values at inference. In particular, according to our experiments and stress database, while by discarding all missing data, as a simplistic yet common approach, no prediction can be made for 34% of the data at inference, our approach can achieve accurate predictions, as high as 78%, for missing samples. Also, our experiments show that the proposed framework obtains a cross-validation accuracy of 86.8% even if more than 50% of samples within the features are missing.

arXiv:2104.12657

tsrobprep - an R package for robust preprocessing of time series data

Data cleaning is a crucial part of every data analysis exercise. Yet, the currently available R packages do not provide fast and robust methods for cleaning and preparation of time series data. The open source package tsrobprep introduces efficient methods for handling missing values and outliers using model based approaches. For data imputation a probabilistic replacement model is proposed, which may consist of autoregressive components and external inputs. For outlier detection a clustering algorithm based on finite mixture modelling is introduced, which considers time series properties in terms of the gradient and the underlying seasonality as features. The procedure allows to return a probability for each observation being outlying data as well as a specific cause for an outlier assignment in terms of the provided feature space. The methods work robust and are fully tunable. Moreover, by providing the auto_data_cleaning function the data preprocessing can be carried out in one cast, without comprehensive tuning and providing suitable results. The primary motivation of the package is the preprocessing of energy system data. We present application for electricity load, wind and solar power data.

arXiv:2104.09677

Large Scale Record Linkage in the Presence of Missing Data

Record linkage is aimed at the accurate and efficient identification of records that represent the same entity within or across disparate databases. It is a fundamental task in data integration and increasingly required for accurate decision making in application domains ranging from health analytics to national security. Traditional record linkage techniques calculate string similarities between quasi-identifying (QID) values, such as the names and addresses of people. Errors, variations, and missing QID values can however lead to low linkage quality because the similarities between records cannot be calculated accurately. To overcome this challenge, we propose a novel technique that can accurately link records even when QID values contain errors or variations, or are missing. We first generate attribute signatures (concatenated QID values) using an Apriori based selection of suitable QID attributes, and then relational signatures that encapsulate relationship information between records. Combined, these signatures can uniquely identify individual records and facilitate fast and high quality linking of very large databases through accurate similarity calculations between records. We evaluate the linkage quality and scalability of our approach using large real-world databases, showing that it can achieve high linkage quality even when the databases being linked contain substantial amounts of missing values and errors.

arXiv:2102.10506공식 코드 · Papers with Code

Practical graph signal sampling with log-linear size scaling

Graph signal sampling is the problem of selecting a subset of representative graph vertices whose values can be used to interpolate missing values on the remaining graph vertices. Optimizing the choice of sampling set using concepts from experiment design can help minimize the effect of noise in the input signal. While many existing sampling set selection methods are computationally intensive because they require an eigendecomposition, existing eigendecompostion-free methods are still much slower than random sampling algorithms for large graphs. In this paper, through optimizing sampling sets towards the D-optimal objective from experiment design, we propose a sampling algorithm that has complexity comparable to random sampling algorithms, while reaching accuracy similar to existing eigendecomposition-free methods for a broad range of graph types.

arXiv:2104.01482

Training Deep Normalizing Flow Models in Highly Incomplete Data Scenarios with Prior Regularization

Deep generative frameworks including GANs and normalizing flow models have proven successful at filling in missing values in partially observed data samples by effectively learning -- either explicitly or implicitly -- complex, high-dimensional statistical distributions. In tasks where the data available for learning is only partially observed, however, their performance decays monotonically as a function of the data missingness rate. In high missing data rate regimes (e.g., 60% and above), it has been observed that state-of-the-art models tend to break down and produce unrealistic and/or semantically inaccurate data. We propose a novel framework to facilitate the learning of data distributions in high paucity scenarios that is inspired by traditional formulations of solutions to ill-posed problems. The proposed framework naturally stems from posing the process of learning from incomplete data as a joint optimization task of the parameters of the model being learned and the missing data values. The method involves enforcing a prior regularization term that seamlessly integrates with objectives used to train explicit and tractable deep generative frameworks such as deep normalizing flow models. We demonstrate via extensive experimental validation that the proposed framework outperforms competing techniques, particularly as the rate of data paucity approaches unity.

arXiv:2103.16336공식 코드 · Papers with Code

Fast model-based clustering of partial records

Partially recorded data are frequently encountered in many applications and usually clustered by first removing incomplete cases or features with missing values, or by imputing missing values, followed by application of a clustering algorithm to the resulting altered dataset. Here, we develop clustering methodology through a model-based approach using the marginal density for the observed values, assuming a finite mixture model of multivariate $t$ distributions. We compare our approximate algorithm to the corresponding full expectation-maximization (EM) approach that considers the missing values in the incomplete data set and makes a missing at random (MAR) assumption, as well as case deletion and imputation methods. Since only the observed values are utilized, our approach is computationally more efficient than imputation or full EM. Simulation studies demonstrate that our approach has favorable recovery of the true cluster partition compared to case deletion and imputation under various missingness mechanisms, and is at least competitive with the full EM approach, even when MAR assumptions are violated. Our methodology is demonstrated on a problem of clustering gamma-ray bursts and is implemented at https://github.com/emilygoren/MixtClust.

arXiv:2103.08323

Spatiotemporal Tensor Completion for Improved Urban Traffic Imputation

Effective management of urban traffic is important for any smart city initiative. Therefore, the quality of the sensory traffic data is of paramount importance. However, like any sensory data, urban traffic data are prone to imperfections leading to missing measurements. In this paper, we focus on inter-region traffic data completion. We model the inter-region traffic as a spatiotemporal tensor that suffers from missing measurements. To recover the missing data, we propose an enhanced CANDECOMP/PARAFAC (CP) completion approach that considers the urban and temporal aspects of the traffic. To derive the urban characteristics, we divide the area of study into regions. Then, for each region, we compute urban feature vectors inspired from biodiversity which are used to compute the urban similarity matrix. To mine the temporal aspect, we first conduct an entropy analysis to determine the most regular time-series. Then, we conduct a joint Fourier and correlation analysis to compute its periodicity and construct the temporal matrix. Both urban and temporal matrices are fed into a modified CP-completion objective function. To solve this objective, we propose an alternating least square approach that operates on the vectorized version of the inputs. We conduct comprehensive comparative study with two evaluation scenarios. In the first one, we simulate random missing values. In the second scenario, we simulate missing values at a given area and time duration. Our results demonstrate that our approach provides effective recovering performance reaching 26% improvement compared to state-of-art CP approaches and 35% compared to state-of-art generative model-based approaches.

arXiv:2103.06845공식 코드 · Papers with Code

A hierarchical Bayesian model to find brain-behaviour associations in incomplete data sets

Canonical Correlation Analysis (CCA) and its regularised versions have been widely used in the neuroimaging community to uncover multivariate associations between two data modalities (e.g., brain imaging and behaviour). However, these methods have inherent limitations: (1) statistical inferences about the associations are often not robust; (2) the associations within each data modality are not modelled; (3) missing values need to be imputed or removed. Group Factor Analysis (GFA) is a hierarchical model that addresses the first two limitations by providing Bayesian inference and modelling modality-specific associations. Here, we propose an extension of GFA that handles missing data, and highlight that GFA can be used as a predictive model. We applied GFA to synthetic and real data consisting of brain connectivity and non-imaging measures from the Human Connectome Project (HCP). In synthetic data, GFA uncovered the underlying shared and specific factors and predicted correctly the non-observed data modalities in complete and incomplete data sets. In the HCP data, we identified four relevant shared factors, capturing associations between mood, alcohol and drug use, cognition, demographics and psychopathological measures and the default mode, frontoparietal control, dorsal and ventral networks and insula, as well as two factors describing associations within brain connectivity. In addition, GFA predicted a set of non-imaging measures from brain connectivity. These findings were consistent in complete and incomplete data sets, and replicated previous findings in the literature. GFA is a promising tool that can be used to uncover associations between and within multiple data modalities in benchmark datasets (such as, HCP), and easily extended to more complex models to solve more challenging tasks.

논문 ID: 3336.34293

Creating cloud-free satellite imagery from image time series with deep learning

Optical satellite images are important for environmental monitoring. Unfortunately, such images are often affected by distortions, such as clouds, shadows, or missing data. This work proposes a deep learning approach for cleaning and imputing satellite images, which could serve as a reliable preprocessing step for spatial and spatio-temporal analyzes. More specifically, a coherent and cloud-free image for a specific target date and region is created based on a sequence of images of that region obtained at previous dates. Our model first extracts information from the previous time steps via a special gating function and then resorts to a modified version of the well-known U-Net architecture to obtain the desired output image. The model uses supplementary data, namely the approximate cloud coverage of input images, the temporal distance to the target time, and a missing data mask for each input time step. During the training phase we condition our model with the targets cloud coverage and missing values (disabled in production), which allows us to use data afflicted by distortion during training and thus does not require pre-selection of distortion-free data. Our experimental evaluation, conducted on data of the Landsat missions, shows that our approach outperforms the commonly utilized approach that resorts to taking the median of cloud-free pixels for a given position. This is especially the case when the quality of the data for the considered period is poor (e.g., lack of cloud free-images during the winter/fall periods). Our deep learning approach allows to improve the utility of the entire Landsat archive, the only existing global medium-resolution free-access satellite archive dating back to the 1970s. It therefore holds scientific and societal potential for future analyses conducted on data from this and other satellite imagery repositories.

arXiv:2103.03045공식 코드 · Papers with Code

Factor-Based Imputation of Missing Values and Covariances in Panel Data of Large Dimensions

Economists are blessed with a wealth of data for analysis, but more often than not, values in some entries of the data matrix are missing. Various methods have been proposed to handle missing observations in a few variables. We exploit the factor structure in panel data of large dimensions. Our \textsc{tall-project} algorithm first estimates the factors from a \textsc{tall} block in which data for all rows are observed, and projections of variable specific length are then used to estimate the factor loadings. A missing value is imputed as the estimated common component which we show is consistent and asymptotically normal without further iteration. Implications for using imputed data in factor augmented regressions are then discussed. To compensate for the downward bias in covariance matrices created by an omitted noise when the data point is not observed, we overlay the imputed data with re-sampled idiosyncratic residuals many times and use the average of the covariances to estimate the parameters of interest. Simulations show that the procedures have desirable finite sample properties.