RESUMO
Several studies have been conducted to classify various real life events but few are in medical fields; particularly about breast recurrence under statistical techniques. To our knowledge, there is no reported comparison of statistical classification accuracy and classifiers' discriminative ability on breast cancer recurrence in presence of imputed missing data. Therefore, this article aims to fill this analysis gap by comparing the performance of binary classifiers (logistic regression, linear and quadratic discriminant analysis) using several datasets resulted from imputation process using various simulation conditions. Our study aids the knowledge about how classifiers' accuracy and discriminative ability in classifying a binary outcome variable are affected by the presence of imputed numerical missing data. We simulated incomplete datasets with 15, 30, 45 and 60% of missingness under Missing At Random (MAR) and Missing Completely At Random (MCAR) mechanisms. Mean imputation, hot deck, k-nearest neighbour, multiple imputations via chained equation, expected-maximisation, and predictive mean matching were used to impute incomplete datasets. For each classifier, correct classification accuracy and area under the Receiver Operating Characteristic (ROC) curves under MAR and MCAR mechanisms were compared. The linear discriminant classifier attained the highest classification accuracy (73.9%) based on mean-imputed data at 45% of missing data under MCAR mechanism. As a classifier, the logistic regression based on predictive mean matching imputed-data yields the greatest areas under ROC curves (0.6418) at 30% missingness while k-nearest neighbour tops the value (0.6428) at 60% of missing data under MCAR mechanism.
RESUMO
BACKGROUND: In Zanzibar, little is known about the arboviral disease vector Aedes aegypti in terms of abundance, spatio-temporal distribution of its larval habitats or factors associated with its proliferation. Effective control of the vector requires knowledge on ecology and habitat characteristics and is currently the only available option for reducing the risk of arboviral epidemics in the island nation of Zanzibar. METHODOLOGY: We conducted entomological surveys in households and surrounding compounds from February to May 2018 in the urban (Mwembemakumbi and Chumbuni) and rural (Chuini and Kama) Shehias (lowest government administrative unit) situated in the Urban-West region of Unguja island, Zanzibar. Larvae and pupae were collected, transported to the insectary, reared to adult, and identified to species level. Characteristics and types of water containers were also recorded on site. Generalized linear mixed models with binomial and negative binomial distributions were applied to determine factors associated with presence of Ae. aegypti immatures (i.e. both larvae and pupae) or pupae, alone and significant predictors of the abundance of immature Ae. aegypti or pupae, respectively. RESULTS: The survey provided evidence of widespread presence and abundance of Ae. aegypti mosquitoes in both urban and rural settings of Unguja Island. Interestingly, rural setting had higher numbers of infested containers, all immatures, and pupae than urban setting. Likewise, higher House and Breteau indices were recorded in rural compared to the urban setting. There was no statistically significant difference in Stegomyia indices between seasons across settings. Plastics, metal containers and car tires were identified as the most productive habitats which collectively produced over 90% of all Ae. aegypti pupae. Water storage, sun exposure, vegetation, and organic matter were significant predictors of the abundance of immature Ae. aegypti. CONCLUSIONS: Widespread presence and abundance of Ae. aegypti were found in rural and urban areas of Unguja, the main island of Zanzibar. Information on productive habitats and predictors of colonization of water containers are important for the development of a routine Aedes surveillance system and targeted control interventions in Zanzibar and similar settings.
