AI/ML/DL

k-nearest neighbors algorithm, KNeighborsClassifier, train_test_split, data preprocessing, etc.

KNeighborsRegressor - Reshaping and model training

KNeighborsRegressor - Evaluation metrics (R2, mae, mse, rmse)

Underfitting 과 Overfitting 의 원인 분석 및 해결방법에 대한 간단한 설명

Hyperparameters (n_neighbors, etc.)

Linear regression 에 대한 설명 및 예시 (limitations of KNeighborsRegressor and linear regression models)

Polynomial regression 에 관한 설명 및 구현

Multiple regression, Feature engineering (PolynomialFeatures), Pandas, etc.

Regularization 및 standardization 의 필요성

L1 regularization (Lasso, alpha, graphical analysis, and lasso.coef_ == 0)

L2 regularization, alpha, optimization of model fit, etc.

k-NN Classification 과 probability 의 한계점 및 logistic regression 의 필요성

Logistic regression and Sigmoid function: Mapping inputs to probabilities

Binary classification (logistic regression, decision function, sigmoid, etc.)

Multiclass classification (OvR, softmax)

Gradient descent (stochastic gradient descent and loss function)

Logistic loss (binary cross entropy, data preprocessing, SGDClassifier, etc.)

Model fitting with epochs and early stopping

Decision tree (DecisionTreeClassifier, gini impurity, pruning, unscaled features, and feature importances)

Machine learning workflow, validation set, cross-validation, etc.

Splitter (StratifiedKFold) and grid search (GridSearchCV)

Pitfalls of grid search and a random search with probability distribution (uniform, randint, etc.)

Structured vs Semi-structured data and ensemble learning

Random forest and Extra tree ensemble models

Gradient boosting algorithm, histogram-based gradient boosting (permutation importance), and others (XGBoost and LightGBM)