Unit 4 Sprint 3: Major Neural Network Architectures

This week we will review several popular feed-forward neural network architectures that are common in commercial applications.

• Module 1: RNNs & LSTMs
  • Objectives:
    1. Describe recurrent neural network architecture
    2. Use an LSTM to generate text based on some input
• Module 2: CNNs
  • Objectives:
    1. Describe convolutions and convolutions within neural networks
    2. Apply pre-trained CNNs to image classification problems
• Module 3: Autoencoders
  • Objectives:
    1. Describe the componenets of an autoencoder
    2. Train an autoencoder
    3. Apply an autoencoder to a basic information retreval problem
• Module 4: LSTMs for Time Series Forecasting
  • Objectives:
    1. Understand how to prepare time series data for model ingestion (sliding train-test split)
    2. Understand how to use a LSTM model for time series forecasting applications
    3. Understand the importance of seasonality and trends in time series applications.

Hello world testing