/iceberg-classifier-challenge

A convolutional neural network classifier that can classify iceberg in images that were taken from a satellite.

Primary LanguageJupyter Notebook

Machine Learning Engineer Nanodegree

Project: Iceberg Classifier Challenge - Machine Learning Capstone Project

Overview

The goal of this project was to create a classifier that can successfully classify icebergs in images that were taken from satellites. The project contains these following steps:

  1. Data Analysis including data exploration, exploratory visualization, feature selection and data preprocessing.
  2. Create a Convolutional Neural Network in Keras that predicts the probability of an image containing iceberg and validate the model.
  3. Results analysis and evaluate the robustness of the model.

Software and Libraries

  • Python 3.6
  • Tensorflow
  • Pandas
  • Numpy
  • Matplotlib
  • Keras
  • Jupyter Notebook

Datasets

The datasets are hosted in a Kaggle competition. Please download data from the link provided.

This following link contains both train(train.json.7z) and test(test.json.7z) data. https://www.kaggle.com/c/statoil-iceberg-classifier-challenge/data

Code

Find the jupyter notebook at project\iceberg_classifier.ipynb

Report

Check the pdf file at report\report.pdf