Portfolio Details

Project Overview

USA Accidents (Kaggle) – Exploratory Data Analysis | Portfolio Points Performed in-depth exploratory data analysis (EDA) on a large-scale US road accidents dataset to uncover spatial, temporal, and environmental patterns influencing accident severity. Analyzed key attributes including accident severity, location coordinates (latitude/longitude), distance impact, time of occurrence, and road infrastructure features (signals, stop signs, roundabouts). Segmented features into categorical, numerical, boolean, and integer types to enable structured analysis and efficient preprocessing. Conducted correlation analysis using heatmaps to identify relationships between accident severity and weather/visibility-related variables (temperature, humidity, wind speed, visibility). Visualized geospatial accident distributions to highlight regional clustering and high-risk zones across the US. Explored temporal patterns (day/night, sunrise/sunset, time-based trends) to understand peak accident periods. Identified data quality aspects such as missing values and feature distributions to guide further preprocessing and modeling decisions. Built reproducible EDA workflows using Python, Pandas, NumPy, Seaborn, and Matplotlib in a Jupyter Notebook environment.

Impact: Generated actionable insights to support traffic safety analysis and data-driven decision-making. Skills Highlighted: Data cleaning, feature engineering, correlation analysis, geospatial visualization, statistical exploration. Tools: Python, Pandas, NumPy, Seaborn, Matplotlib, Jupyter Notebook, Kaggle.