AbdullahImran/Fire-Feature-Analysis

Fire Feature Analysis

This repository contains trained models and generated artifacts from feature extraction, dimensionality reduction, representation learning, and clustering experiments performed as part of the fire detection and severity classification project.

Contents

Autoencoder

autoencoder_model_dimention_reduction.keras — a trained Keras autoencoder used in the dimensionality-reduction workflow.

Autoencoder Latent Representations

autoencoder_latent_representations_dimention_reduction.npy — latent representations generated by the autoencoder.

Autoencoder Losses

autoencoder_losses_tri_classification_dimention_reduction.npy — loss values generated during the autoencoder experiment.

EfficientNet Feature Representations

efficientnet_features_tri_classification_feature_extraction.npy — feature representations extracted using EfficientNet.

Fire Feature Representations

fire_features_effnet_tri_classification_feature_extraction.npy — fire-related feature representations generated during the feature-extraction workflow.

SOM Clustering

som_tri_classification_clustering.npy — output of the Self-Organizing Map (SOM) clustering experiment.

Loading the Autoencoder

import tensorflow as tf

autoencoder = tf.keras.models.load_model(
    "autoencoder_model_dimention_reduction.keras"
)

autoencoder.summary()

Loading the NumPy Artifacts

import numpy as np

latent_representations = np.load(
    "autoencoder_latent_representations_dimention_reduction.npy"
)

losses = np.load(
    "autoencoder_losses_tri_classification_dimention_reduction.npy"
)

som_results = np.load(
    "som_tri_classification_clustering.npy"
)

Analysis Workflow

The artifacts in this repository support experimentation involving:

1. Deep feature extraction
2. Representation learning
3. Dimensionality reduction
4. Latent-space analysis
5. Feature analysis
6. Clustering

Project Context

This repository is part of a larger deep learning project containing:

• binary fire detection models
• fire severity classification models
• datasets
• notebooks
• feature extraction
• dimensionality reduction
• clustering
• recommendation generation

Limitations

The generated NumPy artifacts are dependent on the original data preprocessing and model pipelines.

To correctly interpret these artifacts, the corresponding project code and preprocessing procedures should also be considered.

License

No standardized open-source license has been specified for this repository.

Please refer to the original project and dataset terms before redistribution or commercial use.