Hire me!

If you are looking for person for your Data Science/Machine Learning/Artificial Intelligence project, you are probably looking for me! You can find my LinkedIn profile here, and below you can check my example portfolio.

2017

2016

2014

  • Rpart results per various random data samples - application written in Shiny. Goal of this application was to present possibilities of interactive web site powered by Shiny. Source code for this application is here. I also prepared simple presentation in slidify regarding the same subject which can be accessed here with source code for it here. #r #shiny #slidify #shinyapps #rpubs
  • Danger of weather events in United States - simple analysis done during Reproducible Research course at Coursera.org. This analysis is fully reproducible as it was primary goal of this assignment. As "side effect" this analysis shows that event tornado is most dangerous for human life and health and event flood is most expensive in terms of damaged properties and crops. #r #reproducible-research #knitr #rpubs #markdown
  • Heart disease indicators - plain application of rpart classification tree for determining three most influential attributes which could predict diameter narrowing. #r #machine-learning #reproducible-research #knitr #rpubs
  • Counting Twitter mentions - there is easy way to count Twitter occurrences of an interesting term. I wrote report in which I show how to easily retrieve Tweets which contains interesting term. I also count occurrences in fixed time window. This report also generates data frame which contain actual text from that Tweets, which can be easily used for example in text mining. #r #twitter #reproducible-research #knitr #rpubs
  • Popular routes - based on data released by Uber, I tried to pick most often used routes across San Francisco. First I estimated rectangle area and divided it by "cells" with 10 x 10 m dimension. Then I aggregated data points into cells and removed counts that are repeating (I assumed that such counts represent slowly travelling car). Then I picked top 1% of cells which have highest number of counts. Analysis don't give popular route but rather shows popular spots on map, which could help determine popular routes. #r #geospatial #geolocation #reproducible-research #rpubs
  • Mining Titanic with Python at Kaggle - entropy and information gain example in python based on data about Titanic from Kaggle competition. #python #kaggle #classification #pandas #entropy
  • Publications about quantum cryptography (1, 2, 3) - three publications about testing quantum key distribution systems in various fibre networks. #physics #quantum-key-distribution #linux #data-analysis #science

2013

  • OpenHomeAutomation - currently suspended project about home automation. Simple software written in Python which shows temperature on dynamic web page. Data is received from Arduino with temperature sensor. My blog about this project: openhomeautomation.blogspot.com/. #python #arduino #home-automation #flask #sqlite
  • Arbitrator - script in Python which quotes MtGox and Bitcurex and compares asks and bids for possible arbitrage. #python #api #bitcoin #blessings #json
  • ClavisAnalyzer - scripts in Python which are cleaning and extracting data from raw logs from two devices for quantum key distribution. #python #quantum-key-distribution #matplotlib #pandas #physics
  • Quelle - scripts in R for analysing experiment data and creating quality control charts. #r #qcc #ggplot2 #quantum-key-distribution #physics

2012