#### This course is Mostly for People Who Like Programming, and In Particular People Who Like Programming in Python

Intro to Neural Networks for Natural Language Processing

**What this course is about:**

Intro to Neural Networks for Natural Language Processing. Today’s modern machine learning libraries have thoughtful, high-level APIs that allow *anyone* who is comfortable with code to dive in to construction of real-world Neural Networks(NNs) using Deep Learning. Combining Keras and TensorFlow together allows us to learn about how to use NNs and deep learning while abstracting away from a ton of details and keeping the big picture in mind. Total in-Class Hours: 7,5

Important Note: This is not a neural network or Deep Learning theory class. We absolutely will spend time talking about the theory of NNs because we want to pull back the veil a bit and we want to make sure you complete the course with a conceptual understanding of Neural Networks and Deep Learning. But the majority of our time will be spent writing and reading and debugging code written in the high-level parts of the APIs of Keras and TensorFlow. This will let us learn how to write basic Neural Net related programs.

**Who this program is for:**

This course is for people who 1) want to get a sense of what Deep Learning is all about and 2) who’d like to use the high-level libraries Keras and TensorFlow to write code that implement the basics of Deep Learning on their own laptops or in the cloud via Jupyter Notebooks.

**What is Deep Learning?**

What is the Difference Between a Convolutional Neural Network and a Recurrent Neural Network? Stand On The Shoulders of Giants: With Modern, High Level APIs, Big Data Is *Not* Always Needed For Deep Learning. Let’s Code An Image Recognition and Classification Problem in Keras Using A Convolutional Neural Networks. Let’s Code A Natural Language Processing Problem in Keras Using A Recurrent Neural Networks The Care and Feeding of Neural Networks: Hyperparameters, Optimization Functions, Activation Functions, Fully Connected Layers, and Other Things The Make Neural Networks Go. Using TensorBoard, TensorFlow’s Great Visualization Tool For Neural Networks

**You will learn:**

- What are Neural Networks And Why Does Everyone Suddenly Care About Them?
- What Sorts of Data Science Problems Can Neural Networks Solve?
- State of the Art 2017 versus State of the Art 2012: Exactly How Much Has the World Changed In The Last Five Years?
- Introduction to Keras and TensorFlow: Learn to Code Neural Networks Using Python
- Building Recurrent Neural Networks Using Python and TensorFlow
- The Representation of Words as Vectors
- Putting it All Together : Topic Classification and Sentiment Analysis

**Prerequisites & Preparation:**

- Python Programming 101
- Laptop
- Laptop