Spatio-temporal data mining

Intersection DBSCAN (I-DBSCAN) is a variant of DBSCAN, a clustering algorithm that allows capturing amorphic-shaped clusters by density calculation algorithm. I-DBSCAN allows applying DBSCAN on large datasets by lowering the complexity by applying Leader* algorithm.

Data Mining Algos

Machine Learning Classification & Clustering of Iris flowers

CV Project: SOLO and GRU for Hemostatic Plug Segmentation

Customer segmentation using clustering

Clustering Geolocation Data Intelligently in Python

Code used to perform the bibliographic analysis in the paper 'A survey on Neural Recommender Systems: insights from a bibliographic analysis'

Clustering

This project focuses on predicting customer churn in an e-commerce setting using machine learning techniques.

Clustering Algorithms (KMeans, MeanShift, (Merged KMean and MeanShift) and DBSCAN)

DBSCAN Clustering Algorithm with KDtree C# faster Implementation

An attempt at the network anomaly detection task using manually implemented k-means, spectral clustering and DBSCAN algorithms, with manually implemented evaluation metrics (precision, recall, f1-score and conditional entropy) used to evaluate these algorithms.

Clustering of taxis hot pickup zones in NYC

Python implementation of Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm for unsupervised learning. Identifies clusters of varying shapes and sizes in data, robust to noise. Useful for data exploration and anomaly detection.

Fast implementations of various clustering algorithms, trajectory processing, and binary similarity metrics with Python SWIG bindings for select algorithms.