Multiple Imputation for Missing Data Using Genetic Programming

Tran, Cao Truong; Zhang, Mengjie; Andreae, Peter

doi:10.1145/2739480.2754665

Cited by 28 publications

(20 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [2,4] several algorithms of supervised classification are considered, such as Hotdeck, KNN, and Decision Trees. These methods require a labeled training set of data and they use similarity metrics to define the relations among the elements to classify.…”

Section: Related Workmentioning

confidence: 99%

“…Under this context, this paper analyses the problem of text classification, where an algorithm classifies the observed text values within different categories. Existing literature proposes different machine learning and data mining algorithms to solve classification problems in a supervised way [2,3,4]. Algorithms such as K-Nearest-Neighbors (KNN), where text distance metrics can be used to classify elements, are relevant for documents classification problems [3,5].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Evaluación de distintas técnicas de representación de texto y medidas de distancia de texto usando KNN para clasificación de documentos

Calvo-Valverde¹,

Mena-Arias²

2020

View full text Add to dashboard Cite

Actualmente, los datos textuales constituyen una parte fundamental de las bases de datos de todo el mundo y uno de los mayores desafíos ha sido la extracción de información útil a partir de conjuntos grandes de documentos de texto. La literatura existente sobre métodos para resolver este problema es muy extensa, sin embargo, los métodos estadísticos (que utilizan métricas de similitud sobre vectores de palabras) han mostrado resultados muy favorables en el campo de la minería de texto durante los últimos 25 años. Adicionalmente, otros modelos han surgido como una prometedora alternativa para lograr reducción dimensional e incorporación de la semántica en la clasificación de documentos, tal como el modelado de temas. Este proyecto se enfoca en la evaluación de técnicas de representación y medidas de similitud de texto (Coseno, Jaccard y Kullback-Leibler) usando el algoritmo de Vecinos más Cercanos (KNN por sus siglas en inglés), con el fin de medir la efectividad del modelado de temas para reducción dimensional al clasificar texto. Los resultados muestran que la versión más tradicional del vector de palabras y la similitud Jaccard superaron al resto de las combinaciones en la mayoría de los casos de uso. Sin embargo, el análisis estadístico mostró que no hubo una diferencia significativa entre la exactitud obtenida al usar representaciones generadas por la Asignación de Dirichlet Latente (técnica de modelado de temas más conocida como LDA por sus siglas en inglés), y la obtenida usando técnicas tradicionales de clasificación de texto. LDA logró abstraer miles de palabras en menos de 60 temas para el primer conjunto de pruebas. Experimentos adicionales sugieren que el modelado de temas puede llegar a lograr un mejor rendimiento al ser usado para clasificar textos cortos y al incrementar el número de temas permitidos al momento de generar el modelo.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Evaluación de distintas técnicas de representación de texto y medidas de distancia de texto usando KNN para clasificación de documentos

Calvo-Valverde¹,

Mena-Arias²

2020

View full text Add to dashboard Cite

show abstract

“…GPMI is a Genetic Programming (GP) algorithm for multiple imputations that was proposed by Tran et. al [11]. GPMI uses GP as a non-parametric regression method to build mathematical functions that regress missing values of one feature on other features under the control of prediction capability and classification accuracy during the evaluation process.…”

Section: Recent Related Workmentioning

confidence: 99%

“…Table 3 shows the atoms factors of the used performance metrics: True Positive (TP), False Positive (FP), False Negative (FN), True Negative (TN), Positive (P), and Negative (N) instances [24]. using TP, TN, FP, and FN as given in Equations (9), (10), (11) and (12) respectively [25]. `…”

Section: B Evaluation Metricsmentioning

confidence: 99%

A New Hybrid Genetic and Information Gain Algorithm for Imputing Missing Values in Cancer Genes Datasets

Elzeki¹,

Alrahmawy²,

Elmougy³

2019

IJISA

View full text Add to dashboard Cite

A DNA microarray can represent thousands of genes for studying tumor and genetic diseases in humans. Datasets of DNA microarray normally have missing values, which requires an undeniably crucial process for handling missing values. This paper presents a new algorithm, named EMII, for imputing missing values in medical datasets. EMII algorithm evolutionarily combines Information Gain (IG) and Genetic Algorithm (GA) to mutually generate imputable values. EMII algorithm is column-oriented not instance oriented than other implementation of GA which increases column correlation to the class in the same dataset. EMII algorithm is evaluated for imputing the generated missing values in four cancer gene expression standard medical datasets (Colon, Leukemia, Lung cancer-Michigan, and Prostate) via comparing the truth original complete datasets against the imputed datasets. The analysis of the experimental results reveals that the imputed values generated by EMII were almost the same as the original values besides having the same impact on the applied classifiers due to accuracy as similar as the original complete datasets. EMII has a running time of θ(n2), where n is the total number of columns.

show abstract

“…Results show that the multiple imputation outperforms single imputation such as mean/mode, regression and hot deck imputation methods. [159] proposes GPMI, a multiple imputation method that uses genetic programming as a regression method to estimate missing values. Experiments on eight datasets with six levels of missing values compare GPMI with seven other popular and advanced imputation methods on two measures: the prediction accuracy and the classification accuracy.…”

Section: Imputation For Classification With Incomplete Datamentioning

confidence: 99%

Evolutionary Machine Learning for Classification with Incomplete Data

Tran¹

Self Cite

View full text Add to dashboard Cite

<p>Classification is a major task in machine learning and data mining. Many real-world datasets suffer from the unavoidable issue of missing values. Classification with incomplete data has to be carefully handled because inadequate treatment of missing values will cause large classification errors. Existing most researchers working on classification with incomplete data focused on improving the effectiveness, but did not adequately address the issue of the efficiency of applying the classifiers to classify unseen instances, which is much more important than the act of creating classifiers. A common approach to classification with incomplete data is to use imputation methods to replace missing values with plausible values before building classifiers and classifying unseen instances. This approach provides complete data which can be then used by any classification algorithm, but sophisticated imputation methods are usually computationally intensive, especially for the application process of classification. Another approach to classification with incomplete data is to build a classifier that can directly work with missing values. This approach does not require time for estimating missing values, but it often generates inaccurate and complex classifiers when faced with numerous missing values. A recent approach to classification with incomplete data which also avoids estimating missing values is to build a set of classifiers which then is used to select applicable classifiers for classifying unseen instances. However, this approach is also often inaccurate and takes a long time to find applicable classifiers when faced with numerous missing values. The overall goal of the thesis is to simultaneously improve the effectiveness and efficiency of classification with incomplete data by using evolutionary machine learning techniques for feature selection, clustering, ensemble learning, feature construction and constructing classifiers. The thesis develops approaches for improving imputation for classification with incomplete data by integrating clustering and feature selection with imputation. The approaches improve both the effectiveness and the efficiency of using imputation for classification with incomplete data. The thesis develops wrapper-based feature selection methods to improve input space for classification algorithms that are able to work directly with incomplete data. The methods not only improve the classification accuracy, but also reduce the complexity of classifiers able to work directly with incomplete data. The thesis develops a feature construction method to improve input space for classification algorithms with incomplete data by proposing interval genetic programming-genetic programming with a set of interval functions. The method improves the classification accuracy and reduces the complexity of classifiers. The thesis develops an ensemble approach to classification with incomplete data by integrating imputation, feature selection, and ensemble learning. The results show that the approach is more accurate, and faster than previous common methods for classification with incomplete data. The thesis develops interval genetic programming to directly evolve classifiers for incomplete data. The results show that classifiers generated by interval genetic programming can be more effective and efficient than classifiers generated the combination of imputation and traditional genetic programming. Interval genetic programming is also more effective than common classification algorithms able to work directly with incomplete data. In summary, the thesis develops a range of approaches for simultaneously improving the effectiveness and efficiency of classification with incomplete data by using a range of evolutionary machine learning techniques.</p>

show abstract

Multiple Imputation for Missing Data Using Genetic Programming

Cited by 28 publications

References 29 publications

Evaluación de distintas técnicas de representación de texto y medidas de distancia de texto usando KNN para clasificación de documentos

Evaluación de distintas técnicas de representación de texto y medidas de distancia de texto usando KNN para clasificación de documentos

A New Hybrid Genetic and Information Gain Algorithm for Imputing Missing Values in Cancer Genes Datasets

Evolutionary Machine Learning for Classification with Incomplete Data

Contact Info

Product

Resources

About