Auto-Keras Training in Cecil Street, Singapore

This Auto-Keras training introduces participants to AutoML with Auto-Keras, an open-source deep learning framework built on TensorFlow and Keras. Participants will learn how to automate deep learning model selection, architecture search, and hyperparameter tuning using Auto-Keras, enabling faster and more efficient model development.

Introduction to Auto-Keras & Automated Model Selection

• Module 1: Understanding Auto-Keras & AutoML in Deep Learning
  • What is Auto-Keras? Features & benefits
  • Auto-Keras vs. traditional deep learning
  • Auto-Keras vs. other AutoML frameworks (Auto-sklearn, TPOT, Google AutoML)
• Module 2: Setting Up Auto-Keras
  • Installing Auto-Keras and dependencies
  • Overview of Auto-Keras API and workflow
  • Hands-on: First Auto-Keras model
• Module 3: Auto-Keras for Image Classification
  • Automated model building for image classification
  • Hands-on: Training an Auto-Keras image classification model

Advanced Auto-Keras & Model Deployment

• Module 4: Auto-Keras for Text and Tabular Data
  • Auto-Keras for text classification
  • Auto-Keras for structured/tabular data
  • Hands-on: Building a text classification model
• Module 5: Neural Architecture Search (NAS) & Hyperparameter Tuning
  • Understanding NAS in Auto-Keras
  • Optimizing deep learning architectures
  • Hands-on: Experimenting with NAS and hyperparameter tuning
• Module 6: Model Deployment & Performance Evaluation
  • Exporting and deploying trained models
  • Evaluating model performance and bias detection
  • Hands-on: Deploying an Auto-Keras model in a real-world scenario
• Hands-On Labs