This paper was prepared by a first-year student pursuing Data Science and Business Analytics at the University of Warsaw's Faculty of Economic Sciences. The research was conducted as part of an unsupervised learning class led by Professor Dr. hab. Katarzyna Kopczewska. This paper examines PCA-based image compression and chromatic analysis alongside more advanced unsupervised techniques of computational color transfer, including palette extraction and luminance-aware pixel mapping across artworks. Using RGB channel decomposition, k-means clustering, and statistical distribution matching in perceptual color spaces, the project compressed visual data while identifying and reassigning dominant chromatic structures between paintings. Collectively, the analysis demonstrates how dimensionality reduction and palette-based transfer can model, manipulate, and reinterpret pictorial color information, while also revealing the constraining role of luminance in achieving faithful cross-image color transitions.

Tugas ini dilakukan untuk mengetahui hubungan antar variabel yang ada di dalam dataset Titanic-Dataset.csv. Pada tugas ini hanya menggunakan 4 variabel, yaitu Age, SibSp, Parch, dan Fare. Data yang terdapat missing value pada variabel-variabel tersebut akan dihapus agar analisis lebih akurat. Lalu, hubungan antar variabel dianalisis menggunakan matriks korelasi, matriks varians-kovarians, serta nilai eigen dan vektor eigen.

Este tutorial apresenta um passo a passo para visualizar os valores de honorários advocatícios pagos no âmbito do Poder Executivo Federal, a partir de microdados extraídos do Portal da Transparência (CGU) utilizando o software/linguagem R.