Flight-Data-Analysis

This project focuses on analyzing flight journey data through Exploratory Data Analysis (EDA) and building a Naive Bayes classification model to predict flight arrival status (on-time or delayed).

• Part 1: Perform EDA on the dataset to explore and visualize different features, uncover relationships, and interpret trends.
• Part 2: Train and test machine learning models (Naive Bayes) to classify flights as on-time or late.

---

Dataset Overview

The dataset contains 10 features, including:

• Carrier
• Departure Time
• Destination
• Date
• Flight Number
• Origin
• Day of the Week
• Day of the Month

➡️ View Dataset

Day Encoding

• 1 = Monday, 2 = Tuesday, … , 7 = Sunday
• Day of Month: Numerical values representing each calendar day

Carrier Codes

• CO = Continental
• DH = Atlantic Coast
• DL = Delta
• MQ = American Eagle
• OH = Comair
• RU = Continental Express
• UA = United
• US = USAirways

Destinations

• JFK = Kennedy
• LGA = LaGuardia
• EWR = Newark

Origins

• DCA = Reagan National
• IAD = Dulles
• BWI = Baltimore–Washington Int’l

---

Libraries Used

• Pandas → data handling & transformation
• NumPy → numerical operations
• Matplotlib → visualizations (plots, charts, histograms)
• Seaborn → statistical graphics

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

---

Data Preprocessing

Data cleaning and transformation steps included:

• Handling null values
• Frequency analysis
• Checking for inconsistencies
• Preparing features for statistical analysis and modeling

---

Data Visualization

The dataset is visualized to reveal distributions, trends, and correlations using:

• Pairplot
• Relplot
• Histogram
• Pie Chart
• Barplot
• Strip Plot
• Jointplot

---

Machine Learning Model

Naive Bayes Classification

The Gaussian Naive Bayes algorithm was used for prediction.

Naive Bayes Types:

• Gaussian → continuous data
• Multinomial → discrete data
• Bernoulli → binary data

In this case, the Gaussian model was applied.

from sklearn.preprocessing import LabelEncoder
from sklearn.naive_bayes import GaussianNB
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix, plot_confusion_matrix
from sklearn.metrics import accuracy_score, precision_score, recall_score

x_train, x_test, y_train, y_test = train_test_split(gets, target, test_size=99, random_state=5)
model = GaussianNB()
model.fit(x_train, y_train)
model.score(x_test, y_test)

✅ Accuracy Score: 0.8283

---

Confusion Matrix

The confusion matrix evaluates classification performance by comparing predicted vs actual labels.

• True Positives (TP): Correctly predicted positives
• True Negatives (TN): Correctly predicted negatives
• False Positives (FP): Incorrectly predicted positives
• False Negatives (FN): Incorrectly predicted negatives

---

Project Files

• Jupyter Notebook Code