Welcome to the Dog Breed Prediction project! This repository contains scripts, data, and models to classify dog breeds using machine learning techniques. The project leverages a dataset of dog images and breed labels to train a Machine learning model capable of accurately identifying the breed of a dog from an image.
This project walks through the entire process of building a Machine learning model to classify dog breeds from images. The key steps include data preparation, model building, training, and evaluation.
-
Load the Dataset
Download the dataset from Kaggle.
IMPORTANT: Download the dataset from here.
Size: 750MB Images:10000 -
Load Labels
Load the breed labels from the CSV file, which contains the image ID and corresponding breed. -
Check Breed Count
Analyze the distribution of breeds in the dataset to understand the data balance. -
One-Hot Encoding on Labels
Apply one-hot encoding to the labels to convert breed names into a format suitable for model training. -
Load and Preprocess Images
Load the images, convert them into arrays, and normalize them to ensure consistency in model input. -
Check Data Shape and Size
Verify the shape and size of the input data (X) and labels (Y) to ensure they are correctly formatted. -
Build the Model Architecture
Design the Convolutional neural network architecture using popular Machine learning frameworks. -
Train the Model
Split the data into training and validation sets, and fit the model to the data. Track accuracy and loss metrics during training. -
Evaluate the Model
Assess the model's performance on the validation set by calculating the accuracy score. -
Predict Using the Model
Use the trained model to predict the breed of new dog images and evaluate its performance on unseen data.
- Python 2.7
Make surePython 2.7is installed on your system. - Dependencies
Install the required Python libraries.
Algorithm Comparison We evaluated five different 🤖 machine learning algorithms:
Convolutional Neural Network (CNN): Achieved an impressive accuracy of 97.8150%, making it the top performer.
Support Vector Machine (SVM)
Decision Tree
Random Forest
K-Nearest Neighbors (KNN)
Among these, the CNN outperformed the others in terms of accuracy, demonstrating its effectiveness in handling complex image classification tasks.
Accuracy Details
| Algorithm | Accuracy |
|---|---|
| Convolutional Neural Network (CNN) | 97.8150% |
| Support Vector Machine (SVM) | 79.6680% |
| Decision Tree | 78.3800% |
| Random Forest | 80.5990% |
| K-Nearest Neighbors (KNN) | 78.9230% |
Model Optimization: Continue to fine-tune the CNN model to push the accuracy even higher. More Data: Incorporate additional dog breed images to enhance the model’s capabilities. Deployment: Develop a web app that allows users to upload images and receive breed predictions in real-time.
This project is licensed under the MIT License.
The MIT License is a permissive free software license originating at the Massachusetts Institute of Technology (MIT). It is a simple and easy-to-understand license that places very few restrictions on reuse, making it a popular choice for open-source projects.
Key Points:
- Freedom to Use: You can use the software for any purpose.
- Freedom to Modify: You can modify the software and use it as a base for other projects.
- Freedom to Distribute: You can distribute the original or modified software to others.
- No Warranty: The software is provided "as is", without warranty of any kind.
For the full text of the MIT License, please see the LICENSE file in this repository.
We welcome contributions! Please fork this repository and submit a pull request with your changes.
Thank you for checking out this project! If you have any questions or suggestions, feel free to open an issue or reach out.