Table of contents
Project: Data Visualization of Bird Strikes between 2000 – 2011
What is a bird strike ?
A bird strike is strictly defined as a collision between a bird and an aircraft which is in flight or on a take-off or landing roll. The term is often expanded to cover other wildlife strikes - with bats or ground animals. Bird Strike is common and can be a significant threat to aircraft safety.
What to do ?
Perform Exploratory Data Analysis over the given dataset, to address issues related to bird strikes.
Analysis as follows
There are a lot of questions one can ask based on the problem definition, major ones are
How many bird strikes on year basis?
Under what climatic conditions these strikes are happening?
Derive a sample from data set, which is most concentrated with bird strikes and what do various attributes of dataset correspond to this?
What is the relationship between phase of flight and number of strikes?
How cost is interplaying with other attributes such as phase, altitude, climatic conditions etc?
Were pilots warned?
How altitude is related to number of attacks?
Top 10s for appropriate attributes
.... many more relationships between can be analysed.
But after all these questions what is the end goal?
Find when and where these strikes are happening and cost associated with
Find the sample of data, which is more concentrated with attacks.
Problems faced during EDA
Data Cleaning
Data biased towards one particular class and not uniformaly distributed among all classes
Strategy
Data Cleaning
Data Visualisation
Draw Conclusions
Conclusion
This project, exhausted all my statistical knowledge gather till now, there is no technique remaining for me to apply, I thought of every other possibility to explore the data. Not so easy and not so challenging, everything is in right proportion, great to explore for a beginner.