hirushar / titanic-machine-learning-from-disaster

Titanic: Machine Learning from Disaster

Titanic-Machine-Learning-from-Disaster

This is a solution for kaglle Titanic prediction challange

https://www.kaggle.com/c/titanic Kaggle

• Task is predict who is serviving and who has better servival rate

we have 10692 train data
trian data set contain:

• PassengerId
• Survived
• Pclass
• Name
• Sex
• Age
• SibSp
• Parch
• Ticket
• Fare
• Cabin
• Embarked

SO, Firstly i check how mane servivers are thier and map the servival relationship to get idea about the data

so i get the image with age and class.

After that i create servived relation ship with gender

in that image i found a clear relationship servived with gender. lot of females are sevived according to data.

after that i create hypothis that all female are servived and check with data and i got 78% accuracy.[predict_gender.py]]

i use logistioc regression to get more accurace and i got 79%.

from decision tree model it got 97% accuracy but after normalize it finaly i got 81% accurace.

Used tech:

• python
• MatplotLib
• sklearn
• numpy
• pandas

Thank You!

Hirusha R.