Self-Supervised Bipartite Graph Neural Networks with Missing Value Imputation for Small Tabular Data Predictions

Background . Properly constructed decision support systems (DSS) for modelling and forecasting behaviour of dynamic systems provide a possibility for taking into consideration uncertainties of probabilistic, statistical and structural types what results in higher quality of developed models and estimated forecasts. Objective . To consider general reasons for loosing (missing) data in statistical data analysis; to provide categorization of missing data into several groups, and identify the reasons for missing measurements; to provide stepwise system methodology for uncertainty analysis and selection of data imputation techniques; to give an insight into some popular missing values imputation techniques regarding their possible applications. Methods . To solve the problems mentioned the following methods have been used: data categorization approach from business or practical point of view that is necessary for discovering the reasons for availability of systemic and/or random missing values; the modern systemic methodology was hired for analysis of uncertainty causes and missing values imputation; the decision tree based imputation procedures; EM algorithm and regression model approach to forecasting missing data using forecasting functions. Results . The main results of the study are in categorization of the missing data into groups; development of systemic methodology for analysis of uncertainty causes and missing values imputation; providing an analysis for possibilities of missing values imputation with decision trees, EM algorithm and regression models; development of multistep forecasting functions on the basis of autoregression models; illustration of application of some selected perspective methods for missing data imputation. Conclusions . We proposed the six steps system methodology for data imputation which stresses that selection of correct method for imputation is tightly connected with the step-by-step analysis of the gaps causes and finding an appropriate technique for their imputation. The results of imputation sometimes are rather far from the existing data and should be smoothed or even broken from the sample due to their incorrectness. For such cases it should be proposed a new probabilistic-regression method which allows define parameters of the probability interval for the regression aiming missing data imputation. A series of computing experiments performed with EM algorithm, forecast regression based imputation technique and some other approaches shows that it is possible to reach high quality results regarding correct processing of data with missing values.

(1) Background: Gene-expression data usually contain missing values (MVs). Numerous methods focused on how to estimate MVs have been proposed in the past few years. Recent studies show that those imputation algorithms made little difference in classification. Thus, some scholars believe that how to select the informative genes for downstream classification is more important than how to impute MVs. However, most feature-selection (FS) algorithms need beforehand imputation, and the impact of beforehand MV imputation on downstream FS performance is seldom considered. (2) Method: A modified chi-square test-based FS is introduced for gene-expression data. To deal with the challenge of a small sample size of gene-expression data, a heuristic method called recursive element aggregation is proposed in this study. Our approach can directly handle incomplete data without any imputation methods or missing-data assumptions. The most informative genes can be selected through a threshold. After that, the best-first search strategy is utilized to find optimal feature subsets for classification. (3) Results: We compare our method with several FS algorithms. Evaluation is performed on twelve original incomplete cancer gene-expression datasets. We demonstrate that MV imputation on an incomplete dataset impacts subsequent FS in terms of classification tasks. Through directly conducting FS on incomplete data, our method can avoid potential disturbances on subsequent FS procedures caused by MV imputation. An experiment on small, round blue cell tumor (SRBCT) dataset showed that our method found additional genes besides many common genes with the two compared existing methods.

Network representation learning is receiving increasing attention from scholars. Among them, methods based on graph neural networks have become particularly popular. However, most existing methods currently only focus on networks with a single type of relations. In the real world, networks contain a wealth of diverse information, with multiple types of relationships between nodes. In this paper, we propose a graph neural network-based multiplex network representation learning model (GNMRL). We model nodes within each layer of the multiplex network by aggregating neighbor information. Additionally, since nodes are influenced by different network layers, we integrate node interaction information across network layers. We conducted systematic experiments on four datasets, and the results show that GNMRL outperforms other comparison methods in both link prediction and node classification tasks.

 Electrocardiogram signal analysis is very difficult to classify cardiac arrhythmia using machine learning methods. The ECG datasets normally come with multiple missing values. The reason for the missing values is the faults or distortion. When performing data mining, missing value imputation is the biggest task for data preprocessing. This problem could arise due to incomplete medical datasets if the incomplete missing values and cases were removed from the original database. To produce a good quality dataset for better analyzing the clinical trials, the suitable missing value imputation method is used. In this paper, we explore the different machine-learning techniques for the computed missing value in the electrocardiogram dataset. To estimate the missing imputation values, the collected data contains feature dimensions with their attributes. The experiments to compute the missing values in the dataset are carried out by using the four feature selection methods and imputation methods. The implemented results are shown by combined features using IG (information gain), GA (genetic algorithm) and the different machine learning classifiers such as NB (naïve bayes), KNN (K-nearest neighbor), MLP (Multilayer perception), and RF (Random forest). The GA (genetic algorithm) and IG (information gain) are the best suitable methods for obtaining the results on lower dimensional datasets with RMSE (Root mean square error. It efficiently calculates the best results for missing values. These four classifiers are used to analyze the impact of imputation methods. The best results for missing rate 10% to 40% are obtained by NB that is 0.657, 0.6541, 0.66, 0.657, and 0.657, as computed by RMSE (Root mean Square error). It means that error will efficiently reduced by naïve bayes classifier.

Unsupervised classification of satellite image time series (SITS) is prominent in numerous multitemporal remote sensing applications. However, when optical images are concerned, a missing value reconstruction task becomes pivotal, due to the impact of cloud cover on the input SITS. This task is usually addressed as a pre-processing step before feeding the time series to a clustering model. In this paper, we propose a pixelwise SITS clustering algorithm which integrates missing value imputation jointly with the clustering task. Moreover, the clustering process is designed to jointly perform both representation learning and cluster assignment, enabling the proposed model to simultaneously tackle three duties (missing value imputation, representation learning, and cluster assignment), while being trained in an end-to-end manner. The experimental results show that the proposed model performs well compared to other models that address time series imputation and clustering independently. Furthermore, a visualization analysis suggests that the proposed model learns both the imputation and clustering effectively despite being trained simultaneously.

Missing data imputation for location-based sensor data has attracted much attention in recent years. The state-of-the-art imputation methods based on graph neural networks have a priori assumption that the spatial correlations between sensor locations are static. However, real-world data sets often exhibit dynamic spatial correlations. This paper proposes a novel approach to capturing the dynamics of spatial correlations between geographical locations as a composition of the constant, long-term trends and periodic patterns. To this end, we design a new method called Dynamic Adjacency Matrix Representation (DAMR) that extracts various dynamic patterns of spatial correlations and represents them as adjacency matrices. The adjacency matrices are then aggregated and fed into a well-designed graph representation learning layer for predicting the missing values. Through extensive experiments on six real-world data sets, we demonstrate that DAMR reduces the MAE by up to 19.4% compared with the state-of-the-art methods for the missing value imputation task

BackgroundRecent efforts in the field of nutritional science have allowed the discovery of disease-beating molecules within foods based on the commonality of bioactive food molecules to FDA-approved drugs. The pioneering work in this field used an unsupervised network propagation algorithm to learn the systemic-wide effect on the human interactome of 1962 FDA-approved drugs and a supervised algorithm to predict anticancer therapeutics using the learned representations. Then, a set of bioactive molecules within foods was fed into the model, which predicted molecules with cancer-beating potential.The employed methodology consisted of disjoint unsupervised feature generation and classification tasks, which can result in sub-optimal learned drug representations with respect to the classification task. Additionally, due to the disjoint nature of the tasks, the employed approach proved cumbersome to optimize, requiring testing of thousands of hyperparameter combinations and significant computational resources.To overcome the technical limitations highlighted above, we represent each drug as a graph (human interactome) with its targets as binary node features on the graph and formulate the problem as a graph classification task. To solve this task, inspired by the success of graph neural networks in graph classification problems, we use an end-to-end graph neural network model operating directly on the graphs, which learns drug representations to optimize model performance in the prediction of anticancer therapeutics.ResultsThe proposed model outperforms the baseline approach in the anticancer therapeutic prediction task, achieving an F1 score of 67.99%±2.52% and an AUPR of 73.91%±3.49%. It is also shown that the model is able to capture knowledge of biological pathways to predict anticancer molecules based on the molecules’ effects on cancer-related pathways.ConclusionsWe introduce an end-to-end graph convolutional model to predict cancer-beating molecules within food. The introduced model outperforms the existing baseline approach, and shows interpretability, paving the way to the future of a personalized nutritional science approach allowing the development of nutrition strategies for cancer prevention and/or therapeutics.

In real-world production data, missing values often occur randomly or systematically with various missing patterns. Missing values need to be handled properly to build effective prediction models. This paper presents a novel method based on graph representation and graph neural networks for improving prediction in missing value conditions. To utilize the entire information of a training dataset without direct manipulation, all instances of the dataset are represented as graphs of varying sizes, in which nodes and edges represent the observed input variables and their pairwise relationships. Prediction models learn from the graph representations. These models can make predictions of unknown labels for new instances that have arbitrary missing patterns. The superiority of the proposed method was investigated on seven different product failure prediction tasks from a home appliance manufacturer. The proposed method outperformed all other methods in six of the seven tasks.

Graph Neural Networks (GNNs) are the current state-of-the-art models in learning node representations for many predictive tasks on graphs. Typically, GNNs reuses the same set of model parameters across all nodes in the graph to improve the training efficiency and exploit the translationally-invariant properties in many datasets. However, the parameter sharing scheme prevents GNNs from distinguishing two nodes that are isomorphic and that the translation invariance property may not exhibit in real-world graphs. In this paper, we present Graph Representation Learning with Adaptive Mixtures (GRAM), a novel approach for learning node representations in a graph by introducing multiple independent GNN models and a trainable mixture distribution for each node. GRAM is a general framework that can be readily embedded into existing models for enhancing representation learning. It allows for adaptive node clustering with mixtures and joint optimization of all GNN models simultaneously. To achieve this, we develop an Expectation-Maximization algorithm with stochastic gradient descent in a scalable manner. Specifically, in the E-Step, we approximate the posterior probability distribution of the latent cluster membership based on the k-th independent GNN models in the last iteration. In the M-step, we update all GNN models and the mixture parameters based on the refined latent cluster membership for downstream tasks, e.g., node classification. To fully exploit the graph structural information, we further implement a regularization term on unlabeled nodes, which encourages adjacent nodes to have the same mixture distribution. By jointly optimizing the model in an end-to-end manner, GRAM learns to capture different abstract patterns on graphs for different node clusters, which yields discriminative representation for each node. We evaluate GRAM on five benchmark datasets with extensive experiments. GRAM is demonstrated to consistently boost state-of-the-art GNN variants in node classification tasks.

A massive amount of data is emerging continuously in the era of big data; the missing value is a common yet challenging problem. Missing data or missing value can be described as the data whose values in any given dataset are unknown. It is a recurrent phenomenon and can give biased results from the data to be observed and affect the value of the learning process. Therefore, it is mandatory to use missing value imputation techniques. Missing value imputation provides (MVI) optimal solutions for missing values in a dataset. There are diverse missing value imputation techniques such as statistical method and machine learning methods proposed till date, each having their own significances and flaws. Reasonably, good imputation results may be produced by machine learning MVI approaches but they take greater imputation times than statistical approaches usually. In this paper, we have reviewed some of the missing value imputation approaches proposed. In order to show the efficiency of their proposed approaches, they had compared their proposed methods with some baseline imputation algorithms like mode/median, k-nearest neighbor, Naive Bayes, support vector machine. The outcome of each proposed method is analyzed and their extension scopes from the perspectives of research focus. The main idea of this literature is to give a general review of missing value imputation approaches.

In various fields of information examination, for example, AI, profoundly getting the hang of missing information is a typical issue. Missing qualities should be tended to since they can adversely affect the exactness and adequacy of prescient models. This research investigates how data discretization affects deep learning methods for filling the missing values in datasets with continuous features. They provide a unique method for imputing missing values using deep neural networks (DNNs) called extravagant expectation maximization-deep neural network (EEM-DNN). This approach discretizes continuous features into separate intervals initially. This is justified by treating the issue of missing value imputation as a classification work, with the missing values being considered a distinct class. A DNN, designed explicitly for imputation, is then trained using the discretized data. The expectation maximization concepts are incorporated into the network architecture, and as a result, the network iteratively improves its imputation predictions. They run comprehensive experiments on several datasets from different fields to gauge the efficacy of the suggested strategy. The effectiveness of EEM-DNN is compared to that of other imputation approaches, such as traditional imputation techniques and deep learning methods without data discretization. Our findings show that data discretization significantly enhances imputation accuracy. In terms of imputation accuracy and prediction performance on downstream tasks, the EEM-DNN method regularly performs better than alternative methods. It also examines if various discretization techniques affect the overall imputation process. They find that the trade-off between bias and variance in imputed data depends on the discretization method selected. This highlights the significance of choosing a suitable discretization approach depending on the unique properties of the dataset.

Recently, data-driven decision-making has attracted great interest; this requires high-quality datasets. However, real-world datasets often feature missing values for unknown or intentional reasons, rendering data-driven decision-making inaccurate. If a machine learning model is trained using incomplete datasets with missing values, the inferred results may be biased. In this case, a commonly used technique is the missing value imputation (MVI), which fills missing data with possible values estimated based on observed values. Various data imputation methods using machine learning, statistical inference, and relational database theories have been developed. Among them, conventional machine learning based imputation methods that handle tabular data can deal with only numerical columns or are time-consuming and cumbersome because they create an individualized predictive model for each column. Therefore, we have developed a novel imputational neural network that we term the Denoising Self-Attention Network (DSAN). Our proposed DSAN can deal with tabular datasets containing both numerical and categorical columns; it considers discretized numerical values as categorical values for embedding and self-attention layers. Furthermore, the DSAN learns robust feature expression vectors by combining self-attention and denoising techniques, and can predict multiple, appropriate substituted values simultaneously (via multi-task learning). To verify the validity of the method, we performed data imputation experiments after arbitrarily generating missing values for several real-world tabular datasets. We evaluated both imputational and downstream task performances, and we have seen that the DSAN outperformed the other models, especially in terms of category variable imputation.

Targeted proteomics utilizing antibody-based proximity extension assays provides sensitive and highly specific quantifications of plasma protein levels. Multivariate analysis of this data is hampered by frequent missing values (random or left censored), calling for imputation approaches. While appropriate missing-value imputation methods exist, benchmarks of their performance in targeted proteomics data are lacking. Here, we assessed the performance of two methods for imputation of values missing completely at random, the previously top-benchmarked ‘missForest’ and the recently published ‘GSimp’ method. Evaluation was accomplished by comparing imputed with remeasured relative concentrations of 91 inflammation related circulating proteins in 86 samples from a cohort of 645 patients with venous thromboembolism. The median Pearson correlation between imputed and remeasured protein expression values was 69.0% for missForest and 71.6% for GSimp (p = 5.8e-4). Imputation with missForest resulted in stronger reduction of variance compared to GSimp (median relative variance of 25.3% vs. 68.6%, p = 2.4e-16) and undesired larger bias in downstream analyses. Irrespective of the imputation method used, the 91 imputed proteins revealed large variations in imputation accuracy, driven by differences in signal to noise ratio and information overlap between proteins. In summary, GSimp outperformed missForest, while both methods show good overall imputation accuracy with large variations between proteins.

Missing values in tabular data lakes can severely impact data analysis and diminish the performance in downstream applications. We highlight that a robust imputation strategy should properly take three aspects of variety into consideration: source of imputed value, the types of tables involved, and the data types of the missing value. Existing imputation methods rely on estimation-based approaches (using a model trained on data from the same table to estimate missing values) or search-based approaches (retrieving values from other tables). Unfortunately, none of these approaches effectively incorporate all three aspects of variety. To address this gap, we propose , a novel framework that uses a C ombination of E stimation-based and S earch-based methods for missing value I mputation in D ata lakes. contains three core modules: (1) the , which efficiently discovers candidate values from tables by exploiting the contextual information; (2) the , which introduces an influence function and a sampling-based exploration strategy to yield accurate estimated values; (3) the , which determines the most suitable method based on table-level and column-level statistics. Extensive experiments conducted on three data lakes demonstrate that effectively and efficiently addresses the missing value problem.

Currently, many databases contain missing values, especially in medical data. Statistical and data mining approaches often require complete data conditions, where these two approaches will not provide adequate performance if the data contains missing values. Several techniques have been made to overcome missing values, one of which is by deleting data containing missing values. However, this approach will omit a lot of information if the data found includes many missing values. This study used an imputation approach (filling in the missing attributes) with a clustering approach. One of the most common clustering approaches is K-Means Clustering. In K-means clustering, the value of the centroid gets from the closest observed value. In this study, we propose updating the centroid value based on the harmonic average of the distance across all observations per centroid. This method is known as K-Harmonic Means Clustering (KHM). We proposed a new program approach for a mixed dataset on three scenarios for missing values of 10%, 20%, and 30%. From the experiments conducted on experimental data sets containing missing values, we get a small proportion of missing values (10%) with a small number of clusters or K, which gives a smaller RMSE value compared to other scenarios.