adult income prediction example 240 – 246
exploratory data analysis 241 – 243
AI (artificial intelligence) 3
ANN (artificial neural network) 95
attribution methods, saliency mapping 161 – 162
AUC (area under the curve) 253
BCE (binary cross-entropy) loss function 97
BERT (bidirectional encoder representations from transformers) 215
correcting label bias through reweighting 262 – 266
Diagnostics+ AI system 11 – 12
fairness through unawareness 261 – 262
CART (classification and regression tree) algorithm 34
CBOW (continuous bag of words) 208
CNNs (convolutional neural networks) 52, 95, 130 – 140, 200, 260
probability landscape 140 – 141
training and evaluating 138 – 140
visual attribution methods 147 – 148
coarse-grained activation map 148
Computer Vision Annotation Tool (CVAT) 174
concept detectors, network dissection
continuous bag of words (CBOW) 208
CVAT (Computer Vision Annotation Tool) 174
data leakage, Diagnostics+ AI system 11
data types, PyTorch tensors 286 – 287
DataLoader class 137, 288, 290
datasets, fairness and 266 – 268
breast cancer diagnosis 90 – 91
building 9 – 10, 12 – 14
diabetes progression prediction 24 – 26, 46 – 48
DiCE (diverse counterfactual explanations) 277
distributed representations 207
DNNs (deep neural networks) 52, 95 – 104, 130
training and evaluating 101 – 104
equality of opportunity and odds 251 – 254
counterfactual explanations 275 – 279
adult income prediction example 241 – 243
high school student performance predictor 59 – 63
model-agnostic methods 91 – 95
exponential kernel function 107
FAIR (Facebook’s AI Research) 284
adult income prediction example 240 – 246
exploratory data analysis 241 – 243
datasheets for datasets 266 – 268
equality of opportunity and odds 251 – 254
interpretability and 256 – 261
discrimination via input features 256 – 259
discrimination via representation 260 – 261
correcting label bias through reweighting 262 – 266
fairness through unawareness 261 – 262
FCNNs (fully connected neural networks) 95
feature_extraction function 181
fine-grained activation map 148
fully connected neural networks (FCNNs) 95
GAMs (generalized additive models) 16, 40 – 51, 271
for Diagnostics+ diabetes progression 46 – 48
GDPR (General Data Protection Regulation) 12, 267
global interpretability 16, 74 – 87
partial dependence plots 74 – 80
GPT (generative pretrained transformer) 215
grad-CAM (gradient-weighted class activation mapping) 18, 157 – 160
gradient-based methods, saliency mapping 156 – 157
gradient-boosting algorithm 67
GradientBoostingClassifier class 71
guided backpropagation 153 – 155
high school student performance predictor
exploratory data analysis 59 – 63
ICE (individual conditional expectation) plots 18
IDC (invasive ductal carcinoma) detection 127 – 128
integrated gradients method 148
discrimination via input features 256 – 259
discrimination via representation 260 – 261
global interpretability 74 – 87
partial dependence plots 74 – 80
local interpretability 105 – 115
measuring similarity 217 – 220
principal component analysis 220 – 225
intrinsic interpretability techniques 15
IoU (Intersection over Union) score 177
label bias, correcting through reweighting 262 – 266
limitations of network dissection 198
running network dissection 179 – 183
quantifying alignment 177 – 178
visual understanding 168 – 169
LIMEs (local interpretable model-agnostic explanations) 18, 272
LimeTabularExplainer class 109
local interpretability 16, 105, 147
See also model-agnostic methods
machine learning systems 4 – 9
MAE (mean absolute error) 29, 295
MAPE (mean absolute percentage error) 29
mean squared error (MSE) 34, 296
MLPs (multilayer perceptrons) 95
Model parameter randomization test 161
model-agnostic methods 16 – 125, 147
training and evaluating 101 – 104
exploratory data analysis 91 – 95
global interpretability 74 – 87
partial dependence plots 74 – 80
high school student performance predictor
exploratory data analysis 59 – 63
SHapley Additive exPlanations 115 – 119
model-specific interpretability techniques 15
automatic differentiation 291 – 293
MSE (mean squared error) 34, 296
multilayer perceptrons (MLPs) 95
natural language processing (NLP) 201
quantifying alignment 177 – 178
neural word embeddings 206 – 214
NLP (natural language processing) 201
operations, PyTorch tensors 288
PCA (principal component analysis) 18, 220 – 225, 272
PDPs (partial dependence plots) 18, 74 – 80, 239, 272
perturbation-based methods 147
PIL (Python Imaging Library) 143
post-hoc interpretability techniques 15, 147
adult income prediction example 244 – 246
diabetes progression prediction 24 – 26, 46 – 48
high school student performance predictor 59 – 63
principal component analysis (PCA) 18, 220 – 225, 272
probability landscape, CNNs 140 – 141
automatic differentiation 291 – 293
quantifying alignment, network dissection 177 – 178
RandomForestClassifier class 244
regression splines, GAMs 42 – 46
regulatory noncompliance, Diagnostics+ AI system 12
ReLU (rectified linear unit) 98, 153, 293
reweighting, correcting label bias through 262 – 266
RMSE (root mean squared error) 29
RNNs (recurrent neural networks) 52, 95, 213
ROC (receiver operator characteristic) 253
probability landscape 140 – 141
training and evaluating 138 – 140
visual attribution methods 147 – 148
exploratory data analysis 128 – 130
gradient-based methods 156 – 157
guided backpropagation 153 – 155
guided Grad-CAM technique 157 – 160
vanilla backpropagation 148 – 153
exploratory data analysis 203 – 206
measuring similarity 217 – 220
principal component analysis 220 – 225
neural word embeddings 206 – 214
sentiment analysis 201 – 203, 213 – 214
SHAP (SHapley Additive exPlanations) 18, 115 – 119, 272
Sigmoid activation function 294
SMEs (subject matter experts) 46
SmoothGrad (smooth gradients) 18, 148
SoTA (state-of-the-art) machine learning techniques 134
t-SNE (t-distributed stochastic neighbor embedding) 225 – 230, 261, 272
TF-IDF (term frequency inverse document frequency) 208
torch.nn.Module base class 293
torch.nn.Sequential container 293
torchvision package 137, 170, 284
UAT (user acceptance testing) 10
vanilla backpropagation 148 – 153
visual attribution methods, CNNs 147 – 148
visual understanding, layers and units 168 – 169
VoTT (Visual Object Tagging Tool) 174
weakly model-dependent techniques 148
diabetes progression prediction 24 – 26
diabetes progression prediction 46 – 48
Word2Vec (Word to Vector) 208 – 212