GitHub - rahulOCR/Carnival-Wars-: HackerEarth Machine Learning Challenge: Carnival Wars!

rahulOCR / Carnival-Wars- Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

HackerEarth Machine Learning Challenge: Carnival Wars!

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
ai.py		ai.py
predictions.csv		predictions.csv
readme.txt		readme.txt
sample_submission.csv		sample_submission.csv
test.csv		test.csv
train.csv		train.csv

Repository files navigation

-> source code: ai.py
-> Features used: maximum and minimum price along with charges1 and charges2
-> models used: 
# model = Perceptron()
# model = svm.SVC()
# KNeighborsClassifier(n_neighbors=10)
# model = GaussianNB()

   KNeighborsClassifier is used which gives more closer results with just above features
   While training 60% data is used to train and rest 40% data is used for predictions from train.csv and got 78.2% accuracy
-> Removal of redund rows with '0' in missing coordinates are used in training as well as pridictions are used