Index
A
- acoustic model, speech recognition
- about / Acoustic model
- deep belief networks / Deep belief networks
- recurrent neural networks / Recurrent neural networks
- CTC / CTC
- attention-based models / Attention-based models
- activation function
- activation functions
- about / Network design
- actor-critic methods
- about / Actor-critic methods
- variance reduction, baseline / Baseline for variance reduction
- advantage estimator / Generalized advantage estimator
- Adadelta / Adadelta
- Adagrad / Adagrad
- adaptive learning
- about / Adaptive learning
- rate annealing / Rate annealing
- momentum / Momentum
- Nesterovs acceleration / Nesterov's acceleration
- Newtons method / Newton's method
- Adagrad / Adagrad
- Adadelta / Adadelta
- AI
- training, to master Go / Training AI to master Go, Upper confidence bounds applied to trees
- AlphaGo
- policy gradients / Policy gradients in AlphaGo
- anomaly detection
- about / What is anomaly and outlier detection?
- real-world applications / Real-world applications of anomaly detection
- anomaly detection, techniques
- about / Popular shallow machine learning techniques
- data modeling / Data modeling
- detection modeling / Detection modeling
- deep auto-encoders used / Anomaly detection using deep auto-encoders
- Apache Spark
- about / Sparkling Water
- applications, neural networks
- about / Applications in industry
- signal processing / Signal processing
- medical / Medical
- aautonomous car driving / Autonomous car driving
- business / Business
- pattern recognition / Pattern recognition
- speech production / Speech production
- area under the curve (AUC) / Labeled Data
- Artificial Intelligence (AI)
- about / What is a data product?
- artificial intelligence (AI) / What is machine learning?
- artificial model
- and biological models, differences / Similarities between artificial and biological models
- asynchronous methods
- about / Asynchronous methods
- Atari Breakout
- about / Atari Breakout
- random benchmark / Atari Breakout random benchmark
- screen, preprocessing / Preprocessing the screen
- deep convolutional network, creating / Creating a deep convolutional network
- Q-learning, convergence issue / Convergence issues in Q-learning
- Policy gradients, versus Q-learning / Policy gradients versus Q-learning
- autoencoders
- about / Autoencoders, Summary of autoencoders
- regularization techniques / Regularization techniques for autoencoders
B
- back-propagation algorithm
- about / The back-propagation algorithm
- linear regression / Linear regression
- logistic regression / Logistic regression
- back-propagation / Back-propagation
- back-propagation through time (BPTT) / Character-based model
- boldness
- about / Testing
- Boltzmann machine / Hopfield networks and Boltzmann machines, Boltzmann machine
- Boltzmann Machines (BM) / Deep learning algorithms
C
- character-based models, language modeling
- about / Character-based model
- data, preprocssing / Preprocessing and reading data
- data, reading / Preprocessing and reading data
- LSTM network / LSTM network
- training / Training
- sampling / Sampling
- example training / Example training
- collective anomaly / Data modeling
- Connectionist Temporal Classification (CTC) / CTC
- contextual anomaly / Data modeling
- contractive autoencoders
- about / Contractive autoencoders
- reference link / Contractive autoencoders
- convolutional layers
- about / Intuition and justification, Convolutional layers
- stride and padding / Stride and padding in convolutional layers
- pre-training / Pre-training
- Convolutional Neural Networks (CNN) / Deep learning algorithms
- cross-entropy method / The cross-entropy method
D
- data collection / Steps Involved in machine learning systems
- DataFrame
- about / Sparkling Water
- data modeling
- techniques / Data modeling
- point anomaly / Data modeling
- contextual anomaly / Data modeling
- collective anomaly / Data modeling
- data processing / Steps Involved in machine learning systems
- data product
- about / What is a data product?
- Data Science systems
- reference link / A summary of testing
- decision trees / Decision trees
- deep auto-encoders
- used, for anomaly detection / Anomaly detection using deep auto-encoders
- deep belief networks
- about / Deep belief networks
- Deep Belief Networks (DBN) / Deep learning algorithms
- deep belief networks (DBN) / Deep belief networks
- deep learning / Deep learning
- about / What is deep learning?
- fundamental concepts / Fundamental concepts
- feature learning / Feature learning
- algorithms / Deep learning algorithms
- deep learning, applications
- about / Deep learning applications
- speech recognition / Speech recognition
- object recognition and classification / Object recognition and classification
- deep neural networks (DNN) / Deep belief networks
- denoising autoencoders / Denoising autoencoders
- deployment
- about / Deployment
- POJO model export / POJO model export
- REST API / Anomaly score APIs
- conclusion / A summary of deployment
- detection modeling
- about / Detection modeling
- supervised / Detection modeling
- unsupervised / Detection modeling
- semi-supervised / Detection modeling
- discrete cosine transform (DCT) / Preprocessing
- distributed learning
- via Map/Reduce / Distributed learning via Map/Reduce
- dynamic games
- about / Dynamic games
- experience replay / Experience replay
- Epsilon greedy / Epsilon greedy
E
- early game playing AI
- about / Early game playing AI
- Efficient BackProp
- reference link / Network design
- end-to-end evaluation / End-to-end evaluation
- A/B testing / A/B Testing
- energy based model (EBM)
- about / Restricted Boltzmann machines
- examples, anomaly detection
- about / Examples
- MNIST digit anomaly recognition / MNIST digit anomaly recognition
- Electrocardiogram pulse detection / Electrocardiogram pulse detection
- Excess-Mass (EM) / Unlabeled Data
- explanatory Power
- about / Testing
F
- False Positive Rate (FPR) / Labeled Data
- formal elegance
- about / Testing
- fruitfulness
- about / Testing
G
- Gaussian mixture model (GMM) / Acoustic model
- genetic algorithms
- applying / Applying genetic algorithms to playing games
- GPUs (Graphical Processing Units)
- versus, CPU(Central Processing Unit) / GPU versus CPU
H
- H2O
- about / H2O, Getting started with H2O
- hidden Markov models (HMMs) / Acoustic model
- HOGWILD
- used, for parallel SGD / Parallel SGD using HOGWILD!
- about / Parallel SGD using HOGWILD!
- Hopfield networks / Hopfield networks and Boltzmann machines
- hyper-parameters tuning / Hyper-parameters tuning
I
- Inferior Temporal (IT) / Similarities between artificial and biological models
K
- K-means / K-means
- Kaggle
- Keras / Keras, Sample deep neural net code using Keras
- Key Performance Indicators (KPIs)
- about / Testing
- Kullback-Leibler (KL) / Sparse autoencoders
L
- language modeling
- about / Language modeling
- word-based models / Word-based models
- character-based / Character-based model
- learning policy functions
- policy gradients / Policy gradients for learning policy functions
- Limited Memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) / Newton's method
- linear regression / Linear regression, Linear regression
- logistic regression / Logistic regression
- long short term memory (LSTM) network / Long short term memory
- LSTM (long short-term memory neural network) / Speech recognition
M
- machine learning
- about / What is machine learning?
- approaches / Different machine learning approaches
- supervised learning / Supervised learning
- unsupervised learning / Unsupervised learning
- reinforcement learning / Reinforcement learning
- steps / Steps Involved in machine learning systems
- techniques/algorithms / Brief description of popular techniques/algorithms
- linear regression / Linear regression
- decision trees / Decision trees
- K-means / K-means
- naïve Bayes / Naïve Bayes
- support vector machines / Support vector machines
- cross-entropy method / The cross-entropy method
- neural networks / Neural networks
- deep learning / Deep learning
- real life applications / Applications in real life
- Mass-Volume (MV) / Unlabeled Data
- mean squared error (MSE)
- about / Network design
- Mel Frequency Cepstral Coefficients (MFCC) / Preprocessing
- min-max algorithm
- used, to value game states / Using the min-max algorithm to value game states, Implementing a Python Tic-Tac-Toe game
- MLPClassifier
- Model-based approaches
- about / Model-based approaches
- model-based learning / Actor-critic methods
- model selection
- about / Testing
- model validation
- about / Model validation
- labeled data / Labeled Data
- unlabeled data / Unlabeled Data
- momentum / Momentum
- Monte Carlo Tree Search (MCTS) / Training AI to master Go
- deep learning / Deep learning in Monte Carlo Tree Search
- MSE (mean squared error) / Anomaly detection using deep auto-encoders
- Multi-Layer Perceptrons (MLP) / Deep learning algorithms
N
- N-grams / N-grams
- naïve Bayes / Naïve Bayes
- Nesterovs acceleration / Nesterov's acceleration
- network design
- about / Network design
- Neural language models / Neural language models
- neural networks / Neural networks
- about / Why neural networks?
- characteristics / Fundamentals
- neurons and layers / Neurons and layers
- function xor, code example for / Code example of a neural network for the function xor
- neurons / Neurons and layers
- Newtons method / Newton's method
- Not a Number (NaN) / Backpropagation through time
O
- open source libraries
- about / Popular open source libraries – an introduction
- Theano / Theano
- TensorFlow / TensorFlow
- Keras / Keras
- deep neural net code, Keras used / Sample deep neural net code using Keras
- open source package / A popular open source package
- outlier detection
P
- parallel SGD
- HOGWILD used / Parallel SGD using HOGWILD!
- parsimony
- about / Testing
- Plain Old Java Object (POJO)
- about / POJO model export
- point anomaly / Data modeling
- POJO model export
- about / POJO model export
- policy-based learning / Actor-critic methods
- policy gradients
- for learning policy functions / Policy gradients for learning policy functions
- in AlphaGo / Policy gradients in AlphaGo
- pooling layers / Pooling layers
- Precision-Recall (PR) / Labeled Data
- Principal Component Analysis (PCA)
- about / Autoencoders
- limitations / Autoencoders
- Python Tic-Tac-Toe game
- implementing / Implementing a Python Tic-Tac-Toe game
Q
- Q-function
- about / Q-function
- Q-learning
- in action / Q-learning in action
- Convergence issue / Convergence issues in Q-learning
- versus Policy gradients / Policy gradients versus Q-learning
- Q-Learning
- about / Q-Learning
R
- rate annealing / Rate annealing
- Recurrent neural networks (RNN)
- about / Recurrent neural networks
- one-to-one / Recurrent neural networks
- one-to-many / Recurrent neural networks
- many-to-one / Recurrent neural networks
- many-to-many indirect / Recurrent neural networks
- many-to-many direct / Recurrent neural networks
- implementing / RNN — how to implement and train
- backpropagation, through time algorithm / Backpropagation through time
- gradients, vanishing / Vanishing and exploding gradients
- gradients, exploding / Vanishing and exploding gradients
- Long short term memory (LSTM) / Long short term memory
- regularization techniques
- for autoencoders / Regularization techniques for autoencoders
- denoising autoencoders / Denoising autoencoders
- contractive autoencoders / Contractive autoencoders
- sparse autoencoders / Sparse autoencoders
- reinforcement learning / Reinforcement learning
- Resilient Distributed Data
- about / Sparkling Water
- REST API
- reference link / Anomaly score APIs
- about / Anomaly score APIs
- restricted Boltzmann machines
- about / Restricted Boltzmann machines, Restricted Boltzmann machine
- Hopfield networks / Hopfield networks and Boltzmann machines
- Boltzmann machine / Hopfield networks and Boltzmann machines, Boltzmann machine
- TensorFlow, implementation / Implementation in TensorFlow
- reference link / Implementation in TensorFlow
- deep belief networks / Deep belief networks
- Restricted Boltzmann Machines (RBM) / Deep learning algorithms
- root mean square (RMS) / Adadelta
S
- simplicity
- about / Testing
- Sparkling Water
- about / Sparkling Water
- sparse autoencoders
- about / Sparse autoencoders
- reference link / Sparse autoencoders
- sparsity constraint / Sparse autoencoders
- speech recognition
- about / Speech recognition
- pipeline / Speech recognition pipeline
- speech, as input data / Speech as input data
- preprocessing / Preprocessing
- acoustic model / Acoustic model
- decoding / Decoding
- end-to-end models / End-to-end models
- Stochastic Gradient Descent (SGD)
- about / Parallel SGD using HOGWILD!
- supervised learning / Supervised learning
- supervised learning approach
- support vector machines / Support vector machines
- support vector machines (SVM) / A popular open source package
T
- TensorFlow / TensorFlow
- implementation / Implementation in TensorFlow
- testing
- about / Testing
- model validation / Model validation
- hyper-parameters tuning / Hyper-parameters tuning
- end-to-end evaluation / End-to-end evaluation
- conclusion / A summary of testing
- Theano / Theano
- training
- about / Training
- weights, initialization / Weights initialization
- parallel SGD, HOGWILD used / Parallel SGD using HOGWILD!
- adaptive learning / Adaptive learning
- distributed learning, via Map/Reduce / Distributed learning via Map/Reduce
- Sparkling Water / Sparkling Water
- True Positive Rate (TPR) / Labeled Data
U
- unificatory power
- about / Testing
- unsupervised learning / Unsupervised learning
- upper confidence bounds
- applied, to trees / Upper confidence bounds applied to trees
V
- validation
- about / Testing
- conclusion / Summary of validation
- value-based learning / Actor-critic methods
- value function
- learning / Learning a value function
W
- weights
- initialization / Weights initialization
- word-based models, language modeling
- about / Word-based models
- N-grams / N-grams
- Neural language models / Neural language models