AI-DeepLearning

This repository contains multiple projects showcasing a variety of deep learning models and techniques.

Projects Overview

1: Neural Networks and Deep Learning

• Deep_NN_ImageClassification: Implementing fully connected deep neural networks for image classification.
• Deep_NN_StepByStep: Building and training deep neural networks from scratch.
• Logistic_Regression_NN: Logistic regression applied in a neural network setting.
• Planar_Data_Classification_NN: Binary classification on planar datasets.

2: Improving Deep Neural Networks

• Gradient_Checking: Techniques to verify the correctness of backpropagation.
• Optimization_methods: Various optimization algorithms ((Stochastic) Gradient Descent, Momentum, RMSProp, Adam and Mini-Batches) for training.
• Regularization: Techniques to reduce overfitting in neural networks (Regularization and Dropout).
• TensorFlow_Introduction: Introduction to building models with TensorFlow.

3: Structuring Machine Learning Projects

• ML_Strategy: Guidelines on how to organize and structure ML projects.

4: Convolutional Neural Networks

• AutonomousDriving_CarDetection: Object detection for autonomous driving.
• CNN_ImageClassification: CNNs applied to image classification.
• ConvNet_StepByStep: CNNs built from scratch step by step.
• FaceRecognition: Implementing face recognition systems.
• ImageSegmentation_UNet: Semantic image segmentation using U-Net architecture.
• NeuralStyleTransfer: Transfering artistic style from one image to another.
• ResNets: Building and training a ResNet50 CNN.
• TransferLearning_ImageClassification: Transfer learning using MobileNetV2 for image classification.

5: Sequence Models

• CharacterLevelLanguageModel: Creating a character-level language model.
• Emojify: Creating a model to map sentences to emojis.
• NeuralMachineTranslation_Attention: Neural Machine Translation using attention mechanisms.
• LSTM_MusicGenerator: Generating music using LSTM networks.
• Transformer_NamedEntityRecognition: Applying transformers to NER tasks.
• Transformer_PreProcessing: Pre-processing of raw text for transformer models.
• Transformer_QuestionAnswering: Question answering systems with transformers.
• TriggerWordDetection: Detecting trigger words in audio streams.
• WordVectors_Embeddings: Exploring word embeddings.

Libraries and Frameworks

Projects utilize TensorFlow, Keras, NumPy, and other deep learning tools, demonstrating skills in neural network design, optimization, and advanced architectures like CNNs and RNNs.

For detailed content and weekly breakdowns of projects, refer to the Deep Learning Specialization courses.