This report is about exploratory data analysis and predicting diabetics. The dataset used to make the prediction is the diabetic_data.csv.
The statistical and graphyical exploratory data analysis was done using PySpark.
In order to predict the outcome, eight machine learning algorithms were utilized. Out of these, five were single machine learning algorithms, and three were ensemble learning algorithms. The five single machine learning algorithms include Logistic Regression (LR), Decision Tree (DT), K-Nearest Neighbour (KNN), Naive Bayes (NB), and Multi-Layer Perceptron (MLP) from Artificial Neural networks (ANN). The three ensemble learning algorithms used were Random Forest (RF), Gradient Boosting (GB), and AdaBoost (AB).
Testing Accuracy: Logistic Regression - 57% Decision Tree - 50% K-Nearest Neighbour - 51% Naive Bayes - 14% Artificial Neural Networks - 50% Random Forest - 58% Gradient Boosting - 59% AdaBoost - 58%
F1 score Logistic Regression - 0.52 Decision Tree - 0.49 K-Nearest Neighbour - 0.50 Naive Bayes - 0.08 Artificial Neural Networks - 0.50 Random Forest - 0.54 Gradient Boosting - 0.53 AdaBoost - 0.54