Vessel-Identity-Detector

This project is under my summer internship'24 at Bharat Electronics Limited. as a machine learning trainee

Maritime Data Analysis and MMSI Classification

This repository contains code and resources for analyzing maritime data, performing data augmentation and standardization, training a RandomForestClassifier for MMSI classification, and creating various visualizations.

Project Structure

1. model.ipynb

This notebook focuses on the preprocessing, training, and evaluation of the RandomForestClassifier for MMSI classification.

• Data Loading: Loads the augmented maritime data.
• Preprocessing:
  • One-hot encoding of categorical features (PORT OF ORIGIN, PORT OF DESTINATION, VESSEL TYPE).
  • Conversion of TIMESTAMP to datetime format and extraction of time features (HOUR, DAY_OF_WEEK, MONTH).
  • Standardization of numerical features (LATITUDE, LONGITUDE, COURSE).
• Model Training and Evaluation:
  • Splits the data into training and testing sets.
  • Trains a RandomForestClassifier.
  • Evaluates the model using accuracy score and classification report.
  • Tries different test data split ratios to compare performance.

2. report.ipynb

Data Augmentation and Standardization

This notebook performs data augmentation and standardization to improve model performance.

• Data Loading: Loads the historical AIS data.
• Data Preprocessing:
  • Removes irrelevant columns (DATE TIME (UTC)).
  • Standardizes numerical features (LATITUDE, LONGITUDE, COURSE).
• Data Augmentation:
  • Adds noise to the numerical features to create an augmented dataset.
  • Combines the original and augmented datasets.
• Data Saving: Saves the augmented and standardized dataset for further use.

3. visualization.ipynb

This notebook creates various visualizations to understand the maritime data better.

• Data Loading: Loads the augmented maritime data.
• Data Preprocessing:
  • Converts TIMESTAMP to datetime format and extracts time features.
• Visualizations:
  • Distributions of LATITUDE, LONGITUDE, and COURSE.
  • Count plots for PORT OF ORIGIN, PORT OF DESTINATION, and VESSEL TYPE.
  • Correlation heatmap for numerical features.
  • Geospatial plot of vessel movements using Plotly.
  • Distributions of vessel movements by HOUR, DAY_OF_WEEK, and MONTH.
  • Histograms, scatter plots, and box plots for training and test data.
  • Feature histograms by MMSI for training and test data.

Installation

1. Clone the repository:

git clone https://github.com/yourusername/maritime-data-analysis.git
cd maritime-data-analysis

2. Install the required dependencies:

pip install -r requirements.txt

Usage

1. Data Augmentation and Standardization:

   Open data_augmentation_standardization.ipynb and run all cells to generate the augmented and standardized dataset.

2. Model Training and Evaluation:

   Open model.ipynb and run all cells to preprocess the data, train the RandomForestClassifier, and evaluate its performance.

3. Visualization:

   Open visualization.ipynb and run all cells to create various visualizations to understand the data better.

Contributing

Contributions are welcome! Please open an issue or submit a pull request for any improvements or new features.

---