section 4: DATA MINING (2023 -2024)

DATA MINING - finding patterns in the noise

Database datasets

• Mining Variables (columns)
  • which dimensions have most impact on data pattern?
  • DIMENSION REDUCTION
    • datasets: optidigits.csv, datasets::swiss
    • PCA (principal component analysis)
    • CDA
    • t-SNE
• Mining Cases (rows, observations)
  • looking for similarities in observations
  • CLUSTERING
    • datasets: penguins.rds, datasets::iris
    • HIERARCHICAL clustering
    • k-MEANS clustering
    • DBSCAN
  • CLASSIFICATION
    • datasets: spambase.csv, mlbench::BreastCancer
    • K-NN (k-nearest neighbours)
    • NAIVE BAYES
    • DECISION TREES

List dataset

• looking for behavioural associations: IF someone buys this THEN they likely will buy that
• ASSOCIATION ANALYSIS
  • datasets: arules::groceries, arules::Epub
  • APRIORI
  • ECLAT
  • CDA

Time series datasets

• how data changes over time
• TIME SERIES ANALYSIS
  • datasets: datasets::AirPassengers, datasets::euStockMarkets
  • TIME-SERIES DECOMPOSTION
  • ARIMA
  • MLP

Text datasets

• SENTIMENT ANALYSIS
  • datasets: import TheIlliad.txt, LittleWomen.txt
  • methods:
    • classsification: BINARY classification
    • comparison: SENTIMENT SCORING, VISUAL WORD PAIRS