AI and Machine Learning for Coders

Index

A

• accuracy
  • for dropout-enabled LSTMs, Using dropout
  • impact of reduced learning rate with stacked LSTMS, Optimizing stacked LSTMs
  • improving with hyperparameter tuning, Tuning the Learning Rate
  • lower for test data than training data, Training the Neural Network
  • with lower learning rate, Adjusting the learning rate
  • for LSTM over 30 epochs, Creating a Text Classifier with RNNs
  • measuring prediction accuracy, Measuring Prediction Accuracy
  • reduced dense architecture results, Exploring the model architecture
  • reporting network accuracy, The Complete Code
  • for stacked LSTM architecture, Stacking LSTMs
  • training accuracy versus validation accuracy, Building a Sarcasm Detector Using Embeddings
• activation functions
  • definition of term, Designing the Neural Network
  • rectified linear unit (relu), Designing the Neural Network, Coding Convolutions
  • sigmoid activation function, CNN Architecture for Horses or Humans
  • softmax activation function, Designing the Neural Network, Multiclass Classification
• adam optimizer, The Complete Code
• AI winter, Why I Wrote This Book
• Android apps
  • creating TensorFlow Lite apps
    • step 1: create new Android project, Step 1. Create a New Android Project
    • step 2: edit layout file, Step 2. Edit Your Layout File
    • step 3: add dependencies, Step 3. Add the TensorFlow Lite Dependencies
    • step 4: add TensorFlow Lite model, Step 4. Add Your TensorFlow Lite Model
    • step 5: write activity code for inference, Step 5. Write the Activity Code to Use TensorFlow Lite for Inference
  • IDE for developing, What Is Android Studio? ...