Hi, I’m Bulent Buyuk. I work as a Planning Strategy and Process Development Specialist at LC Waikiki. This position is a Business Development position for Merchandise Planning processes and aims to improve processes based on data. That’s why I aim to improve myself in data cleaning and analysis and use it in my work.
My data interests; Clustering, Time Series and Forecasting.
UseR! 2019 Toulouse - Talk Forecasting - Eran Raviv
This video is about ForecastComb, a package R where multiple forecasting can be combined. The presenter is Eran Raviv.
The presentation consists of the following 3 parts;
On the ForecastComb package
Other things combinations
How the ForecastComb came to be
The Package’s aim is to provide researchers and practitioners with a comprehensive implementation of the most common ways in which forecasts can be combined. The package in its current version covers 15 popular estimation methods for creating a combined forecasts – including simple methods, regression-based methods, and eigenvector-based methods. It also includes useful tools to deal with common challenges of forecast combination (e.g., missing values in component forecasts, or multicollinearity), and to rationalize and visualize the combination results.
Many widely used statistical techniques such as the following implicitly use forecast combination:
Areas of application are as follows:
This article at r-blogger.com gives an overview of K-Means Clustering, one of the unsupervised learning methods, and shows how to apply it in R.
This article at mode.com introduces the prophet library, developed by the core data science team at facebook, and describes how to make time series forecasting in R with this library.
This article at r-statistics.co contains detailed information with time series analysis and shows how to apply it at R.
The details of the time series are: components of the time series, stationary, de-composition, de-seasonality, de-trend and auto correlation.