Sequential Data Modelling with Keras

Introduction

Sequential data is the more prevalent data form such as text, speech, music, DNA sequence, video, and drawing. Analysing sequential data is one of the key goals of machine learning such as document classification, time series forecasting, sentimental analysis, language translation. In this course, we will teach Sequential data analysis using Keras. Keras is a popular high level programming framework for deep learning. It simplifies the process of building deep learning applications. Instead of coding in low level TensorFlow and provide all the details.

Course Objectives:

  • Recap of RNN and LSTM
  • 1D Convolution
  • Sequence 2 Sequence Model
  • Attention Mechanism

Duration

7 hours, 1 Day Course

Mode of Delivery

Classroom-based, Instructor-led Training

Course Outline

  1. Word Embedding
    1. One Hot Encoding of Words
    2. Word Embedding
    3. Pre-Trained Word Embedding
  2. RNN and LSTM
    1. Recurrent Neural Network (RNN)
    2. Long Short Term Memory (LSTM) and GRU
    3. Stacked RNN
    4. Bidirectional RNN
    5. Case Studies on Time Series Prediction with LSTM
  3. 1D Convolution
    1. 1D Convolution on Sequential Data
    2. Combining 1D Convolution and RNN
  4. Sequence To Sequence Model
    1. What is Seq2Seq Model
    2. Attention Mechanism