Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

‎Machine Learning vs.

Classical Econometrics

‎1. MACHINE LEARNING AND DATA ‎ upervised, Unsupervised, and


S
‎PREPARATION ‎Reinforcement Learning Models

‎Standardization
‎Data Preparation and Cleaning
‎Normalization

‎Principal Components Analysis

‎K-means algorithm
‎2. PRINCIPAL COMPONENTS ANALYSIS AND
‎K-MEANS CLUSTERING
‎centroids

‎K-Means Clustering ‎Euclidean distance

‎Manhattan distance

‎silhouette coefficient

‎Underfitting and Overfitting ‎bias- variance tradeoff


‎3. METHODS OF PREDICTION AND SAMPLE
‎1. MODELS ‎SPLITTING ‎ raining, Validation, and Test Data Sub-
T
‎cross-validation
‎Samples

‎States (S)

‎Actions (A)

‎Rewards (R)

‎exploitation

‎Reinforcement Learning ‎exploration

‎Monte Carlo method

‎temporal difference learning method

‎Deep reinforcement learning

‎ . REINFORCEMENT LEARNING AND


4
‎neural networks
‎NATURAL LANGUAGE PROCESSING

‎text mining

‎tokenized

‎stopwords

‎Natural Language Processing ‎Stemming


‎Lemmatization
‎MACHINE LEARNING
‎N-grams

‎bag of words

‎Categorial Variables ‎one-hot encoding

‎hyperparameter
‎Ridge Regression (L2 regularization)
‎shrinkage penalty
1‎ . CATEGORIAL VARIABLES,
‎REGULARIZATION, AND LOGISTIC ‎Regularization
‎LASSO
‎REGRESSION

‎elastic net

‎logistic function

‎maximum likelihood method

‎Logistic Regression ‎log- likelihood function

‎mean squared forecast error (MSFE)

‎mean absolute forecast error

‎classification and regression trees

‎Decision Trees
‎Entropy
‎Information gain
‎Gini measure
‎2.PREDICTION
‎2.DECISION TREES, ENSEMBLE LEARNING, K- ‎Bootstrap aggregation (or bagging)
‎NEAREST NEIGHBORS, AND SUPPORT
‎VECTOR MACHINES ‎Ensemble Learning ‎Random forests

‎Boosting

‎ -Nearest Neighbors and Support Vector


K
‎Machines

‎artificial neural networks (ANNs)

‎activation function
‎Neural Networks
‎gradient descent algorithm

‎learning rate

‎3. NEURAL NETWORKS AND PERFORMANCE ‎1. True positives (TP)

‎2. False positives (FP)


‎confusion matrix
‎3. True negatives (TN)

‎Predictive Performance
‎4. False negatives (FN)

‎receiver operating characteristic (ROC) curve ‎area under the ROC curve (AUC)

You might also like