Image Classification using Neural Networks (CIFAR-10)

This project explores the development and evaluation of neural network models for image classification using the CIFAR-10 dataset. The project utilizes TensorFlow and Keras to build and train different network architectures.

Project Highlights:

• Data Loading and Exploration: Loading and visualizing the CIFAR-10 dataset, including class distribution.
• Data Preprocessing: Scaling image data and applying one-hot encoding to target variables.
• Model Development:
  • Implementation of a basic Deep Neural Network (DNN).
  • Implementation of a Deep and Wide Neural Network to improve classification performance.
  • Utilization of techniques like Batch Normalization and Dropout.
• Model Training and Evaluation:
  • Training models with different optimizers and learning rates.
  • Implementing Early Stopping to prevent overfitting.
  • Utilizing Data Augmentation for improved model robustness.
  • Evaluating models based on accuracy, loss, precision, and recall.
• Results and Comparison: Comparing the performance of the developed models.

Technologies Used:

• Python
• TensorFlow
• Keras
• NumPy
• Matplotlib

Dataset:

The project uses the CIFAR-10 dataset, which consists of 60,000 32x32 color images in 10 classes, with 6,000 images per class. There are 50,000 training images and 10,000 test images.

Getting Started:

1. Clone the repository.
2. Install the required libraries (tensorflow, keras, numpy, matplotlib).
3. Run the notebook to reproduce the results.

Future Enhancements:

• Experiment with more advanced CNN architectures.
• Explore transfer learning techniques.
• Hyperparameter tuning for optimal model performance.