Deep Learning [Certificate]

A collection of my study notes and projects in the course Deep Learning (taught by Andrew Ng).

X-mind:

Course 1 - Neural Network & Deep Learning

Course 2 - Improve Deep Neural Network

Course 3 - Structuring Machine Learning Projects

Course 4 - Convolutional Neural Networks

Course 5 - Sequence Models

Study Notes:

1. Logistic regression in Python

2. Build a deep neural network

3. Initialization, regularization, gradient checking

4. Optimization methods: mini-batch GD, Momentum, Adam

5. TensorFlow tutorial

6. Build a Convolutional Neural Network

7. Keras tutorial

8. Build a Recurrent Neural Network

9. Word embeddings and debiasing

Projects

1. Planar data classification with one didden layer

Implement a binary classification neural network with one hidden layer. This model can classify red points and blue points in the same image with accuracy more than 90%.

2. A deep neural network classifying cats

Build a deep neural network to classifie cats vs. non-cats images with accuracy more than 80%.

3. A Classifier recognizing smiling faces

Implement a classifier in Keras. This model can recognize people's faces and classify them as "happy" or "not happy" with accuracy more than 90%.

4. A ResNet recognizing numbers in sign

Implement a ResNet in Keras for a classification problem. This model can recognize signs representing numbers from 0 to 5 with accuracy more than 85%.

5. Car detection with YOLO algorithm

Use YOLO model to detect cars and their positions in images.

6. Neural style transfer

Implement the Neural Style Transfer algorithm and generate novel artistic images using this algorithm.

7. Face recognition

Build a face recognition system to identify a person.

8. Character level language model - Dinosaurus island

Implement a character level language model which can give names to dinosaurs with cool endings like 'saurus', 'don' etc.

9. Improvise a jazz solo with an LSTM network

Implement a model that uses an LSTM to generate music.

10. Emojify

Build an emojifier which can add the most appropriate emoji to the end of a sentence. For example:

• Input sentence: Congratulations on the promotion!
• Output sentence: Congratulations on the promotion! 👍

11. Neural machine translation on dates

Use an attention model to translate human-readable dates ("25th of June, 2009") into machine-readable dates ("2009-6-25").

12. Trigger word detection

Trigger word detection is the technology that allows devices like Amazon Alexa, Google Home, Apple Siri, and Baidu DuerOS to wake up upon hearing a certain word. For this project, our trigger word will be "Activate". Every time it hears you say "activate", it will make a "chiming" sound.