Publications by Maciej Kuchciak
Market Basket Analysis of Online Retail Data Using FP-Growth
Introduction Significance of FP-Growth Methodology Selection of Dataset Step by Step FP-Growth Application of FP-Growth Hyperparameter Tuning in FP-Growth Stability and Variability of FP-Growth Patterns Technical Insights Efficiency and Performance of FP-Growth Challenges in Scalability and Optimization Results Conclusion Introduction ...
3383 sym R (3103 sym/13 pcs)
T-SNE: A Dimensionality Reduction Technique Exploration
Introduction Overview of dimensionality reduction in unsupervised learning Significance of t-SNE Methodology Selection of dataset Step by step t-SNE Application of t-SNE Hyperparameter Tuning in t-SNE Technical Insights Distance Preservation in t-SNE Understanding the Cost Function: Kullback–Leibler Divergence Scaling and Optimization C...
26473 sym R (2822 sym/6 pcs) 6 img
DBSCAN: Density-Based Spatial Clustrering of Applications with Noise, technique exploration
Abstract Introduction Overview of Clustering in Unsupervised Learning Significance of DBSCAN Methodology Understanding DBSCAN Core Concepts Reachability and Connectivity Visualization of DBSCAN Selection of Synthetic Dataset Results and Analysis Exploration of eps and minPts Effects of Varying eps Influence of minPts on cluster formati...
13125 sym R (2479 sym/5 pcs) 13 img
What are the most important steps in basket analysis with association rules?
Introduction Basket analysis, often referred to as market basket analysis, is a methodology designed to uncover associations between items within extensive datasets, such as transactional records in a retail environment. Primarily utilized in association rule learning, this technique aids in comprehending and predicting customer purchasing beh...
3724 sym
Factors to consider in sensitivity analysis of Multidimensional Scaling (MDS) results
Multidimensional Scaling (MDS) is a powerful tool for visualizing the perceived similarities or dissimilarities among items. However, the robustness of MDS outcomes can vary significantly based on methodological choices. This paper explores the critical factors that influence the sensitivity of MDS results, including distance metrics, data pre...
3560 sym
Hierarchical clustering – pros, cons, interpretation, application
Introduction Hierarchical clustering is a method of cluster analysis that targets to create a hierarchy of clusters. Unlike partitioning methods like k-means, where the number of clusters must be specified in advance, hierarchical clustering builds nested clusters by either merging or splitting them successively. This approach can be very usefu...
3778 sym 1 img
Overview of available packages for clustering methods – classes, results, switching possibilities
Introduction Clustering, a fundamental technique in data analysis and machine learning, groups similar objects based on their characteristics. This paper aims to provide overview of different clustering methods available in R, focusing on their classes, results, flexibility, strengths and weaknesses, and the ease with which analysts can switch ...
3153 sym