Used same methodology for both question.The state diagram is attached on output folder.

Merged all the dataset based on Id number and then converted into .csv format.

Created data frame with necessary features.

Then labeled the dataset by whom the output is dependable.

Converted data frame into scalar form.

Then trained the test and train datset and the test data is 0.20 of actual data.

Applied KNN Classifier and predicted outcome by using confusion matrix.Calculated the accuracy of the results.

Take an input from the customer. Predicted the outcome by corelating the attributes and confusion matrix. Then calculated the mean error rate.

Tools:pandas,numpy,operator,sklearnmetrics(confusionmatrixandaccuracyscore),matplotlib,KNN-Classifier,sklearn model selection