AIR CRASH ANALYTICS

Description

Performed an exploratory data analysis on the Aircrash dataset obtained from Kaggle. Visualized the findings using Tableau to showcase meaningful insights from the data.

Role:

Data Analyst

Technology
Introduction

I performed an EDA on the dataset to obtain some meaningful interpretations from the data. I visualized the following using Tableau:

Total Crashes in the world

I analyzed the dataset to find out the total crashes over time. I then visualized the result on the map of the world to highlight the areas where the air crashes have taken place.

output
EDA to find fatalities vs year

To understand the casualties taking place per year, I did some exploratory data analysis to find the relation. I grouped them into 3 categories-
1. People Onboard
2. Onboard fatalities
3. Ground fatalities. Below is the result that I found

output
Reasons for crash

Based on the description of the crash in the dataset, I created an infographic to identify the most common reasons for an aircrash. These results were obtained by Text mining the "description of aircrash" column in the dataset. Displayed below are my findings

output
Text Mining

Based on the description of the crash in the dataset, I did a bit of text mining to find the most frequent words with their most correlated words. For example "runway" has a pretty high correlation with "overran" which could mean that the airplane overran the runway-leading to the crash.

output
Top Locations of crash

I tried to determine the most common areas where air crashes are taking place. This statistic would help the countries listed at the top. They can then focus more on minimizing the air crashes in their country by investing in better training of pilots and using top-notch airplane parts.

output
Flight with most crashes

To identify the type of flights with the most crashes, I visualized the frequency of the crashes and mapped it with the flight type. This result will certainly help the companies at the top of the graph. It will help them realise that their products are faulty and are leading to air crashes.

output
Reflection

I believe that this data analysis certainly carries a value and meaning. Based on the results obtained, businesses can get a reality check of their products and further improve upon them to prevent air crashes in the future.