Enhances construction site safety using YOLO for object detection, identifying hazards like workers without helmets or safety vests, and proximity to machinery or vehicles. HDBSCAN clusters safety cone coordinates to create monitored zones. Post-processing algorithms improve detection accuracy.

Person ReID

Document-level semantic clustering. Unsupervised topic modelling.

The thesis presents the parallelisation of a state-of-the art clustering algorithm, FISHDBC. This objective has been achived by improving the main data structures and components of the algorithm: HNSW, MST and HDBSCAN. My contribution is based on a lock-free strategy, completely wrote in Python.

Lightning-fast image color clustering with C-based RGB localization/euclidean distance calculation. Supports DBSCAN/HDBSCAN, Shapely geometry.

Using TFIDF, UMAP, and HDBSCAN to analyze product reviews

HDBSCAN Tuning for BERTopic Models

Core Spanning Graph published in ICDE 2022

Implementation of statistics algorithms for Machine Learning & Data Mining. The algorithms were implemented with the Scikit-Learn Library

EIGEN FREQUENCY CLUSTERING USING [KMEANS] [KMEANS & PCA ] [DBSCAN] [HDBSCAN]

Solución al reto BBVA Contigo, Hack BBVA 2021

Optimize clustering labels using Silhouette Score.

Defines a boundary around cluster centers in a given point-layer shapefile.

NeuralMap is a data analysis tool based on Self-Organizing Maps

Data Science algorithms for Qlik implemented as a Python Server Side Extension (SSE).

Library and hand-made clustering algorithms are implemented in this project

Making word clouds more interesting

My solution for Kaggle NYC Taxi Fare Prediction ( ranked 21st/1463)

High Energy Physics particle tracking in CERN detectors

NLP on Korean news articles. Automatic topic extraction through dynamic clustering.