Pytorch implementation of the paper 'Gaussian Mixture Proposals with Pull-Push Learning Scheme to Capture Diverse Events for Weakly Supervised Temporal Video Grounding' (AAAI2024).

Learning second order dynamical system

Python library for adaptive Gaussian mixture state estimation. Useful for navigation and tracking in nonlinear non-Gaussian systems. Capable of incorporating negative information and other imprecise evidence.

80/100 Artificial Intelligence, Clustering and Classification

Implemented an auto-clustering tool with seed and number of clusters finder. Optimizing algorithms: Silhouette, Elbow. Clustering algorithms: k-Means, Bisecting k-Means, Gaussian Mixture. Module includes micro-macro pivoting, and dashboards displaying radius, centroids, and inertia of clusters. Used: Python, Pyspark, Matplotlib, Spark MLlib.

SVM with Gaussian mixture uncertainty

Algorithms for checking the accuracy of a clustering result with known classes, computing cluster validity indices, and generating plots for comparing them. The package is compatible with K-means, fuzzy C means, EM clustering, and hierarchical clustering (single, average, and complete linkage).

Unsupervised learning algorithms to cluster students of a public school

Machine learning assignment on heart disease prediction using PCA, feature selection, K-Means, GMM and hierarchical clustering.

A novel image compressor based on a mixed integer linear program

A minimal working example of the spectral mixture kernel

Machine learning (BGMM/GP) code for the Catassembly Triad framework in JACS. Validates catalyst efficacy prediction via triad descriptors (attachability, controllability, detachability).

Interactive visualization of reverse vs forward KL divergence for multi-modal distribution fitting (mode-seeking vs mode-covering). Zero-dependency, browser-only.

Classical ML algorithms from scratch in NumPy: linear/logistic regression, k-means, PCA, RBF kernel SVM (SMO), and Gaussian mixture EM, each unit-tested against scikit-learn