Machine Learning Models

This repository showcases a collection of machine learning models built with Python, Scikit-learn, TensorFlow, and Keras. The models span various machine learning tasks, including classification, regression, and image classification, offering hands-on examples of applying different algorithms and techniques.

Projects

1. Tweet Emotions Classification

• Objective: Classify tweet emotions (e.g., joy, sadness, surprise).
• Model: Dense Neural Network.
• Preprocessing: Tokenization, padding, and text normalization.
• Accuracy: ~85% on the test set.
• Technologies: Python, TensorFlow, Keras.
• Link: Tweet Emotions Classification

2. BBC Sports News Classification

• Objective: Classify BBC sports news articles into categories such as football, rugby, and cricket.
• Model: Dense Neural Network.
• Preprocessing: Tokenization, stopword removal, and text vectorization.
• Accuracy: ~90% on the test set.
• Technologies: Python, Scikit-learn, Keras.
• Link: BBC Sports News Classification

3. Weather Classification (Computer Vision)

• Objective: Classify weather images (cloudy, rainy, sunny) using Convolutional Neural Networks (CNN).
• Model: CNN with data augmentation.
• Accuracy: ~95% on the validation set.
• Technologies: Python, TensorFlow, Keras.
• Link: Weather Classification (Computer Vision)

4. E-commerce Product Classification

• Objective: Classify e-commerce product descriptions into categories like Electronics, Household, Books, and Clothing.
• Model: Long Short-Term Memory (LSTM) network.
• Preprocessing: Text vectorization using GloVe and FastText embeddings.
• Accuracy: ~92% on the test set.
• Technologies: Python, Keras, TensorFlow.
• Link: E-commerce Product Classification

⚙️ Tech Highlights

• Languages & Tools: Python, Jupyter Notebook
• Frameworks: Scikit-learn, TensorFlow, Keras
• NLP Techniques: Tokenization, padding, stopword removal, GloVe & FastText embeddings
• Computer Vision: CNN architectures with data augmentation
• Model Architectures: Dense Neural Networks, LSTMs
• Evaluation Metrics: Accuracy, Confusion Matrix, Validation Scores

🛠 Technologies Used

• Python – Core programming language for all models and data handling
• Scikit-learn – For traditional machine learning algorithms and pipelines
• TensorFlow – Deep learning framework for building and training models
• Keras – High-level API for creating and training neural networks
• Pandas – For data manipulation and preprocessing
• NumPy – For numerical operations and array manipulation
• Matplotlib – For visualization of training metrics and results

Installation

To run any of the models:

1. Clone the repository:

   git clone https://github.com/AmaedaQ/Machine-Learning-Models.git
   cd Machine-Learning-Models

2. Install the required dependencies:

   pip install -r requirements.txt

3. Run the notebook for the specific project you'd like to explore.