In this blog we will be doing a project based on image classification where our problem statement describe us to classifies the images into two categories i.e. Emergency & Non-Emergency vehicle which is a binary classification problem and we will be solving using neural network.

Before diving deep into this project I would recommend you all to please go through the basics of working with images in my deep learning blog. I am sure you will like it and then proceed ahead with exciting theory and practical’s coming ahead.

Coming to emergency vehicles there can various types like police cars, ambulance or even fire trucks.

Source: Analytics Vidhya (Emergency Vehicle)

Any other vehicle which does not belong to police cars, ambulance and firetrucks come under the category of Non-Emergency vehicle.

Source: Analytics Vidhya (Non-Emergency Vehicle)

STEPS TO SOLVE PROBELM.

1. Loading the data.
2. Creating training, validation and test data: We will train on train data and evaluate on validation data and test on test data.
3. Defining the architecture of our model: We will define how many input neurons, nos. of neurons in hidden layer and output layer.
4. Compile the model (loss function / optimizers): Here we will be defining our cost function which will be used during backward propagation.
5. _Train the model: _We will train our model and define nos. of epochs.
6. Evaluate models performance on training and validation data.

**VIEWING DATASET: **We have dataset in a form of csv file and below we see the first five rows of it. The first column represent the image name and the corresponding column will tell us whether the image is an emergency vehicle or not.

#deep-learning #machine-learning #analytics-vidhya #knowledge-sharing