Machine Learning Model for Imbalanced Cholera Dataset in Tanzania

Abstract

Cholera epidemic remains a public threat throughout history, affecting vulnerable population living with unreliable water and substandard sanitary conditions. Various studies have observed that the occurrence of cholera has strong linkage with environmental factors such as climate change and geographical location. Climate change has been strongly linked to the seasonal occurrence and widespread of cholera through the creation of weather patterns that favor the disease’s transmission, infection, and the growth of Vibrio cholerae, which cause the disease. Over the past decades, there have been great achievements in developing epidemic models for the proper prediction of cholera. However, the integration of weather variables and use of machine learning techniques have not been explicitly deployed in modeling cholera epidemics in Tanzania due to the challenges that come with its datasets such as imbalanced data and missing information. This paper explores the use of machine learning techniques to model cholera epidemics with linkage to seasonal weather changes while overcoming the data imbalance problem. Adaptive Synthetic Sampling Approach (ADASYN) and Principal Component Analysis (PCA) were used to the restore sampling balance and dimensional of the dataset. In addition, sensitivity, specificity, and balanced-accuracy metrics were used to evaluate the performance of the seven models. Based on the results of the Wilcoxon sign-rank test and features of the models, XGBoost classifier was selected to be the best model for the study. Overall results improved our understanding of the significant roles of machine learning strategies in health-care data. However, the study could not be treated as a time series problem due to the data collection bias. The study recommends a review of health-care systems in order to facilitate quality data collection and deployment of machine learning techniques.

Cholera Risk: A Machine Learning Approach Applied to Essential Climate Variables

“…This is an ensemble learning method that fits multiple decision tree classifiers and uses bootstrapping to output a decision averaged over these predictors, to improve prediction accuracy and reduce effects of over-fitting. This approach is consistent with previous literature investigating similar outbreak risk indexes which have demonstrated RF classifiers to be an appropriate method for the application to epidemiological data along with analyses of environmental factors, including for prediction of outbreaks of cholera disease (e.g., [ 35 ]), vector-borne diseases such as dengue (e.g., [ 52 , 53 ]) and malaria (e.g., [ 54 ]), and the avian influenza H5N1 zoonotic disease outbreaks (e.g., [ 55 ]).…”

“…A monthly resolution allows for possible incubation periods of V. cholerae and accommodates potential lags between infection incidence and reporting. This resolution has also been used in previous studies, which have reported that the cholera-incidence response range to environmental variables within a week is generally insignificant [ 35 ]. The chlorophyll-a concentration dataset was converted to a logarithmic scale to accommodate the large range of values, spanning three orders of magnitude.…”

“…As such, the model accuracy would largely reflect the underlying class distribution. Dataset imbalances have previously been identified as a key challenge accompanying machine learning [ 57 ], including applications for cholera analysis [ 35 ]. Previous studies have attempted to overcome the issue of imbalanced datasets through under-sampling (decreasing samples of majority classes) or over-sampling (increasing samples of minority class).…”

“…Using machine learning can provide a solution to these issues as the algorithms are capable of developing and recognizing patterns and complex relationships across large datasets, and this capacity offers particular value when these relationships are either unknown or suspected. To our knowledge, only a few previous studies have explored the use of machine learning techniques for cholera, in terms of serological markers [ 34 ] or environmental drivers [ 35 ].…”

Comparative Analysis of Machine Learning Classifiers for the Prediction of Malaria Incidence Attributed to Climatic Factors

“…Multilayer Perceptron (MPL) and J48 classifier techniques in the Weka tool recently being used in successfully predicting malaria incidents [32][33][34]. Researchers around the globe also used it for prediction of dengue [35][36][37][38] and other public health issues such as Cholera [39], diabetes [40][41][42], heart diseases [43,44].…”