Classifying Galaxies Using Convolutional Neural Networks
graph TD;
A[Input Data] --> B[Train-Test Split]
B --> C[ImageDataGenerator]
C --> D[Model Architecture]
D --> E[Model Compilation]
E --> F[Model Training]
F --> G[Visualize Activations]
G --> H[Final Model]
- "Input Data" represents the initial dataset of galaxy images.
- "Train-Test Split" represents the step where the dataset is split into training and testing sets using the
train_test_splitfunction from the scikit-learn library. - "ImageDataGenerator" represents the usage of the
ImageDataGeneratorclass from TensorFlow to preprocess the input data, which includes rescaling the pixel values to be between 0 and 1. - "Model Architecture" represents the architecture of the convolutional neural network (CNN) model, which consists of two convolutional layers followed by max-pooling layers, a flatten layer, and two fully connected layers with ReLU and softmax activations, respectively. This is defined using the
tf.keras.Sequentialclass. - "Model Compilation" represents the compilation of the CNN model using the Adam optimizer with a learning rate of 0.001, Categorical Crossentropy loss function, and two evaluation metrics: Categorical Accuracy and AUC (Area Under the Curve). This is done using the
compilemethod of the model. - "Model Training" represents the training of the CNN model on the training data using the
fitmethod of the model. The training is done for 8 epochs and the validation data is used to evaluate the model performance. - "Visualize Activations" represents the visualization of the activations of the CNN layers on the test data using the
visualize_activationsfunction from thevisualizemodule. - "Final Model" represents the trained CNN model that can be used for classifying galaxies based on their image data.