gabrielcc2/data_playground
Some studies on databases.
Discovered public repositories for gabrielcc2 in the GitHub catalog.
Some studies on databases.
Just some early experiments on different ideas for supporting code retrieval. Part of an upcoming student project to be carried out at OvGU-Magdeburg, under the supervision of Tatiana Gossen.
Some ideas for a parallel radix join operator, using pthreads and testing a branching approach to task-based parallelization. This was developed as a Scientific Team Project from the DBSE workgroup at the University of Magdeburg, during the Summer Semester 2014.
In this repository are some files from a study on how to support cross-language recommendations for an existing research paper recommender system. We focused on German, English and Spanish. Latent semantic indexing and a framework for automatic translation (enriched with query reformulation and semantic query expansion) were studied. This was a project carried out at the University of Magdeburg Oct 2014-Feb 2015.
Just a homework from a python class
Simple desktop app for indexing and querying text and html files
Image feature extraction, indexing and query by example using LIRE and Lucene. Several approaches can be selected. Submission for programming assignment from Multimedia Retrieval Course
Programming exercises in CUDA from Heterogeneous Parallel Programming Course
Proposed solutions for programming assignments in Octave for Coursera's Machine Learning course: Linear & Logistic regression, neural networks for classification...
A bot for playing a 5x5x5 game of tic-tac-toe, using a very naive approach, but showing that common-sense heuristics can have a good impact on the performance. A similar agent based on reinforcement learning will be shortly added. Submission for Machine Learning course@OvGU
A simple implementation of the ID3 algorithm for learning decision trees. A JSwing and MXGraph-based GUI for visualizing the growth of the tree and the execution of reduced error pruning is also included. Submission for assignment of Machine Learning course.
Some homeworks implementing classifiers, specifically tailored for the car evaluation data set (https://archive.ics.uci.edu/ml/datasets/Car+Evaluation): K-nearest neighbours, Kmeans and NaiveBayes approaches. Submission for a Machine Learning course.
Desktop aplication that can crawl the Web for code snippets, beggining with a set of seed URLs and complying mostly with the robots.txt standards. It uses Lucene to index the webpages and support queries over the crawled data. Different indexes can be created, loaded + stored, allowing for specialized searches. This was developed as a mini-project from an Information Retrieval course, WiSe2014-2015@OvGU.
Here are some simple tests using Lightweight Modular Staging (http://scala-lms.github.io/) as a code generating scheme for producing C code implementing different and combinable optimizations (variants) of the Scan operation for a database. Branch removal, thread parallelization, vectorization using SSE instructions and loop unrolling are included in this release. GPU parallelization has also been tested and will be included in the future. Project carried out at the University of Magdeburg Oct 2014-Apr 2015
HW 3 skeleton for doing BDD with RottenPotatoes. From Engineering Software as a Service using Ruby-on-Rails, a MOOC from UC BerkeleyX.