post-hoc-analysis

Simple, compact, and hackable post-hoc deep OOD detection for already trained tensorflow or pytorch image classifiers.

Critical difference diagrams with Python and Tikz

This repository is created for storing the components of Statistical Tests of One Pop, Two Pops and Three or more pops using Python.

code for "Spectral Removal of Guarded Attribute Information"

Plot critical difference diagrams in Julia

code for "Erasure of Unaligned Attributes from Neural Representations"

about statistical techniques for Data Science

Perform a STEP by STEP multiple mean comparison analysis on R

A collection of INF2178H course projects showcasing the full Data Science lifecycle, including experimental planning, data cleaning, exploratory analysis, modeling, and evaluation. Projects apply qualitative and quantitative methods for knowledge discovery and decision-making.

Parallel evolutionary algorithm that represents images with polygons using edge detection and denoising. Configuration is carried out through normality tests and pairwise comparisons.

A set of statistical methods conducted on a strict set of algorithm's performance readings, utilizing Python

ASIS is a web application developed for the compilation of impact report PDF documents for the tutorial programme (A-STEP) at the University of the Free State.

This project evaluates an A/B/n experiment conducted for the company Eniac to optimize the Click-Through Rate (CTR) on the homepage of its website. The experiment focused on testing two key design variables and was carried out as part of a practical data analytics project during a bootcamp at WBS Coding School.

Industrial tool life (Standzeit) optimization. Statistical comparison of 4 production variants using non-parametric Kruskal-Wallis and Post-hoc tests in R.

A question-driven exploration of a Climate Change Reddit Posts dataset through sentiment analysis, topic finding, hierarchical clustering, and statistics.

This projects aim was to identify which crew members were most likely to survive the Titanic wreck based on features such as Work department, Gender, Age, etc. We use Machine Learning concepts such as Logistic Regression and feature selection as well as data preprocessing.

Gain hands-on experience with ANOVA analysis, understanding its assumptions, and applying it to real-world datasets to understand differences among group means.

A/B testing analysis of Eniac’s homepage buttons using Python, with click-through rates, chi-square significance testing, and post-hoc analysis.

An R project that investigates and visualizes the effect of sex and education on an individual's income level through the use of a full-factorial two-way ANOVA test and relevant post hoc significance tests conducted over a 2014 Pew Research Center dataset consisting of higher education attainment, gender, and income data.