A collection of beginner-friendly deep learning notebooks covering core concepts, architectures, and practical applications using TensorFlow/Keras, NumPy, and Python.
This lab is designed to help you learn by doing — each notebook builds on the previous one with clear theory, code, and hands-on workflow.
📄 01_DL_Packages.ipynb
Introduction to essential deep learning libraries: NumPy, TensorFlow, Keras, data loading, tensors, and basic operations.
📄 02_FeedForward_NN_Keras_TensorFlow.ipynb
Build and train a fully connected neural network from scratch, including:
- Forward pass
- Activation functions
- Loss functions
- Backpropagation (conceptual overview)
- Training loop in Keras
📄 03_Image_Classification_CNN.ipynb
Learn convolutional neural networks (CNNs) for image classification:
- Convolution & feature extraction
- Pooling
- Building a CNN in Keras
- Training on a real image dataset
📄 04_Autoencoder_Anomaly_Detection.ipynb
Unsupervised learning with autoencoders:
- Encoder–decoder architecture
- Latent space
- Reconstruction error
- Detecting anomalies using reconstruction loss
📄 05_CBOW_NLP_Model.ipynb
Hands-on NLP project using word embeddings:
- Tokenization
- Context windowing
- Embedding layers
- Training a CBOW model to predict context words
📄 06_Object_Detection_TransferLearning.ipynb
Introduction to object detection using transfer learning:
- Pre-trained CNN backbones
- Feature extraction
- Fine-tuning
- Bounding box prediction workflow
- Open the notebooks using Jupyter Notebook, VSCode, or Jupyter Lab.
- Work through each notebook in order — they’re structured to guide your learning step-by-step.
Unnecessary system files (like .DS_Store) are removed and added to .gitignore to keep the repository clean and professional.