Search Results for "matlab text classification"

...Code written so as to expose and comment on mathematical steps. The repository includes clustering, regression, classification, neural networks, anomaly detection, and other standard ML topics. Does not rely heavily on specialized toolboxes or library shortcuts.

The Exclusively Dark (ExDARK) dataset is one of the largest curated collections of real-world low-light images designed to support research in computer vision tasks under challenging lighting conditions. It contains 7,363 images captured across ten different low-light scenarios, ranging from extremely dark environments to twilight. Each image is annotated with both image-level labels and object-level bounding boxes for 12 object categories, making it suitable for detection and classification...

This repository hosts the Omniglot dataset for one-shot learning, containing handwritten characters across multiple alphabets along with stroke data. It includes both MATLAB and Python starter scripts (e.g. demo.m, demo.py) to illustrate how to load the images and stroke sequences and run baseline experiments (such as classification by modified Hausdorff distance). The dataset provides both an image representation of each character and the time-ordered stroke coordinates ([x, y, t]) for each instance. ...

The following code is a parallel kNN implementation that uses GPUs for the high dimensional data in text classification. You can use it to classify documents using kNN or to generate meta-features based on the distances between a query document and its k nearest neigbors

We present BudgetedSVM, a C++ toolbox containing highly optimized implementations of three recently proposed algorithms for scalable training of Support Vector Machine (SVM) approximators: Adaptive Multi-hyperplane Machines (AMM), Budgeted Stochastic Gradient Descent (BSGD), and Low-rank Linearization SVM (LLSVM). BudgetedSVM trains models with accuracy comparable to LibSVM in time comparable to LibLinear, as it allows solving highly non-linear classification problems with millions of...

An information extraction library implementing modern algorithms for the extraction of named entities from text.