Search results

The primary aim of this study is to review the studies from different dimensions including type of methods, experimentation setup and evaluation metrics used in the novel approaches proposed for data imputation, particularly in the machine learning (ML) area. This ultimately provides an understanding about how well the proposed framework is evaluated and what type and ratio of missingness are addressed in the proposals. The review questions in this study are (1) what are the ML-based imputation methods studied and proposed during 2010–2020? (2) How the experimentation setup, characteristics of data sets and missingness are employed in these studies? (3) What metrics were used for the evaluation of imputation method?

The review process went through the standard identification, screening and selection process. The initial search on electronic databases for missing value imputation (MVI) based on ML algorithms returned a large number of papers totaling at 2,883. Most of the papers at this stage were not exactly an MVI technique relevant to this study. The literature reviews are first scanned in the title for relevancy, and 306 literature reviews were identified as appropriate. Upon reviewing the abstract text, 151 literature reviews that are not eligible for this study are dropped. This resulted in 155 research papers suitable for full-text review. From this, 117 papers are used in assessment of the review questions.

This study shows that clustering- and instance-based algorithms are the most proposed MVI methods. Percentage of correct prediction (PCP) and root mean square error (RMSE) are most used evaluation metrics in these studies. For experimentation, majority of the studies sourced the data sets from publicly available data set repositories. A common approach is that the complete data set is set as baseline to evaluate the effectiveness of imputation on the test data sets with artificially induced missingness. The data set size and missingness ratio varied across the experimentations, while missing datatype and mechanism are pertaining to the capability of imputation. Computational expense is a concern, and experimentation using large data sets appears to be a challenge.

It is understood from the review that there is no single universal solution to missing data problem. Variants of ML approaches work well with the missingness based on the characteristics of the data set. Most of the methods reviewed lack generalization with regard to applicability. Another concern related to applicability is the complexity of the formulation and implementation of the algorithm. Imputations based on k-nearest neighbors (kNN) and clustering algorithms which are simple and easy to implement make it popular across various domains.

The objective of this research is to examine, validate, and recommend techniques for handling the problem of missingness in observational data. We use a rich observational data set, the Nielsen HomeScan data set, which allows us to effectively combine elements from simulated data sets: large numbers of observations, large number of data sets and variables, allowing elements of “design” that typically come with simulated data, and its observational nature. We created random 20% and 50% uniform missingness in our data sets and employed several widely used methods of single imputation, such as mean, regression, and stochastic regression imputations, and multiple imputation methods to fill in the data gaps. We compared these methods by measuring the error of predicting the missing values and the parameter estimates from the subsequent regression analysis using the imputed values. We also compared coverage or the percentages of intervals that covered the true parameter in both cases. Based on our results, the method of single regression or conditional mean imputation provided the best predictions of the missing price values with 28.34 and 28.59 mean absolute percent errors in 20% and 50% missingness settings, respectively. The imputation from conditional distribution method had the best rate of coverage. The parameter estimates based on data sets imputed by conditional mean method were consistently unbiased and had the smallest standard deviations. The multiple imputation methods had the best coverage of both the parameter estimates and predictions of the dependent variable.

The purpose of this paper is to introduce a new multiple imputation method that can effectively manage missing values in online review data, thereby allowing the online review analysis to yield valid results by using all available data.

This study develops a missing data method based on the multivariate imputation chained equation to generate imputed values for online reviews. Sentiment analysis is used to incorporate customers’ textual opinions as the auxiliary information in the imputation procedures. To check the validity of the proposed imputation method, the authors apply this method to missing values of sub-ratings on hotel attributes in both the simulated and real Honolulu hotel review data sets. The estimation results are compared to those of different missing data techniques, namely, listwise deletion and conventional multiple imputation which does not consider text reviews.

The findings from the simulation analysis show that the imputation method of the authors produces more efficient and less biased estimates compared to the other two missing data techniques when text reviews are possibly associated with the rating scores and response mechanism. When applying the imputation method to the real hotel review data, the findings show that the text sentiment-based propensity score can effectively explain the missingness of sub-ratings on hotel attributes, and the imputation method considering those propensity scores has better estimation results than the other techniques as in the simulation analysis.

This study extends multiple imputation to online data considering its spontaneous and unstructured nature. This new method helps make the fuller use of the observed online data while avoiding potential missing problems.

The purpose of this paper is to focus on the case where – by design – one needs to impute cross-country cross-survey (CCCS) data (situation typical for example among multinational firms who are confronted with the need to carry out comparative marketing surveys with respondents located in several countries). Importantly, while some work demonstrates approaches for single-item direct measures, no prior research has examined the common situation in international marketing where the researcher needs to use multi-item scales of latent constructs. The paper presents problem areas related to the choices international marketers have to make when doing cross-country/cross-survey research and provides guidance for future research.

Multi-country sample of real data is used as an example of cross-sample imputation (292 New Zealand exporters and 302 Finnish ones) the international entrepreneurial orientation (IEO) data. Three variations of the input data are tested: first, imputation based on all the data available for the measurement model; second, imputation based on the set of items based on the invariance structure of the joint items shared across the two groups; and third, imputation based both on examination of the invariance structures of the joint items and the performance of the measurement model in the group where the full data was originally available.

Based on distribution comparisons imputation for New Zealand after completing the measurement model with Finnish data (Model C) gave the most promising results. Consequently, using knowledge on between country measurement qualities may improve the imputation results, but this benefit comes with a downside since it simultaneously reduces the amount of data used for imputation. None of the imputation models leads to the same statistical inferences about covariances between latent constructs than as the original full data, however.

Considering multiple imputation, the present exploratory study suggests that there are several concerns and issues that should be taken into account when planning CCCSs (or split questionnaire or sub-sampling designs). Even if there are several advantages available for well-implemented CCCS designs such as shorter questionnaires and improved response rates, these concerns lead us to question the appropriateness of the CCCS approach in general, due to the need to impute across the samples.

The combination of cross-country and cross-survey approaches is novel to international marketing, and it is not known how the different procedures utilized in imputation affect the results and their validity and reliability. The authors demonstrate the consequences of the various imputation strategy choices taken by using a real example of a two-country sample. The exploration may have significant implications to international marketing researchers and the paper offers stimulus for further research in the area.

Centrifugal compressors are integral components in oil industry, thus effective maintenance is required. Condition-based maintenance and prognostics and health management (CBM/PHM) have been gaining popularity. CBM/PHM can also be performed remotely leading to e-maintenance. Its success depends on the quality of the data used for analysis and decision making. A major issue associated with it is the missing data. Their presence may compromise the information within a set, causing bias or misleading results. Addressing this matter is crucial. The purpose of this paper is to review and compare the most widely used imputation techniques in a case study using condition monitoring measurements from an operational industrial centrifugal compressor.

Brief overview and comparison of most widely used imputation techniques using a complete set with artificial missing values. They were tested regarding the effects of the amount, the location within the set and the variable containing the missing values.

Univariate and multivariate imputation techniques were compared, with the latter offering the smallest error levels. They seemed unaffected by the amount or location of the missing data although they were affected by the variable containing them.

During the analysis, it was assumed that at any time only one variable contained missing data. Further research is still required to address this point.

This study can serve as a guide for selecting the appropriate imputation method for missing values in centrifugal compressor condition monitoring data.

In real-world decision-making, high accuracy data analysis is essential in a ubiquitous environment. However, we encounter missing data while collecting user-related data information because of various privacy concerns on account of a user. This paper aims to deal with incomplete data for fuzzy model identification, a new method of parameter estimation of a Takagi–Sugeno model in the presence of missing features.

In this work, authors proposed a three-fold approach for fuzzy model identification in which imputation-based linear interpolation technique is used to estimate missing features of the data, and then fuzzy c-means clustering is used for determining optimal number of rules and for the determination of parameters of membership functions of the fuzzy model. Finally, the optimization of the all antecedent and consequent parameters along with the width of the antecedent (Gaussian) membership function is done by gradient descent algorithm based on the minimization of root mean square error.

The proposed method is tested on two well-known simulation examples as well as on a real data set, and the performance is compared with some traditional methods. The result analysis and statistical analysis show that the proposed model has achieved a considerable improvement in accuracy in the presence of varying degree of data incompleteness.

The proposed method works well for fuzzy model identification method, a new method of parameter estimation of a Takagi–Sugeno model in the presence of missing features with varying degree of missing data as compared to some well-known methods.

The actual data mining process deals significantly with prediction, estimation, classification, pattern recognition and the development of association rules. Therefore, the significance of the analysis depends heavily on the accuracy of the database and on the chosen sample data to be used for model training and testing. Data mining is based upon searching the concatenation of multiple databases that usually contain some amount of missing data along with a variable percentage of inaccurate data, pollution, outliers and noise. The issue of missing data must be addressed since ignoring this problem can introduce bias into the models being evaluated and lead to inaccurate data mining conclusions. The objective of this research is to address the impact of missing data on the data mining process.

The state of Mato Grosso represents the largest producer and exporter of soybeans in Brazil; given this importance, it was aimed to propose to use the univariate imputation tool for time series, through applications of splines interpolations, in 46 of its municipalities that had missing data in the variables soybean production in thousand tons, production value and soy derivatives in R$ thousand, and also to assess the differences between the observed series and those with imputed values, in each of these municipalities, in these variables.

The proposed methodology was based on the use of the univariate imputation method through the application of cubic spline interpolation in each of the 46 municipalities, for each of the 3 variables. Then, for each municipality, the original series were compared with each observed series plus the values imputed in these variables by the Quenouille test of correlation of time series.

It was observed that, after imputation, all series were compared with those observed and are equal by the Queinouille test in the 46 municipalities analyzed, and the Wilcoxon test also showed equality for the accumulated total of the three variables involved with the production of soybeans. And there were increases of 5.92%, 3.58% and 2.84% for soy production, soy production value and soy derivatives value accumulated in the state after imputation in the 46 municipalities.

The present research and its results facilitate the process of estimates and monitoring the total soy production in the state of Mato Grosso and its municipalities from 1990 to 2018.