Machine Learning

Supervised Learning : Predictive modeling, linear algebra, classification; binary, mulitclass, regression:continous values, knn, logistic regression

feature vector ; qualitative : categorical data - ordinal data ; nominal data - one hot encoding ;

quantitative : numerical; continous, discrete

Unsupervised Learning: structure/representing data

Reinforcement Learning

Labeled data

model training

python Packages

scikit learn

stats model

matplotlib

numpy : numerical computation

scipy

sklearn

pandas

seaborn

exploratory data analysis

model training

precision and recall

loss and validation set

model evaluation procedures: train and test on entire dataset; evaluation metrics;

classification and prediction accuracy

training accuracy; overfit; underfit

decision boundary

train test split ; random state with 4 ; train, test, validation dataset

ramdom sampler

training and testing acuracy

feature selection

confusion matrix

model selection feature selection (cross validation for parameter tuning)

model complexity

predictions of out of sample data

k fold cross validation

ROC Curve and AUC

parameter tuning

optimal tuning parameters

bias and variance

cross valu score

scikit learn grid search and randomize search cross validate one hot encoding

models

linear regression; intercept and coefficient

knn

bayes classification

MLOPS

References: https://www.pinecone.io/learn/k-nearest-neighbor/