After working on data platforms mainly focusing on business intelligence, data warehouse development and data engineering areas in several banks, I have been working as data engineer at Hopi the Shopping App since 2017. I have had roles in the core areas of data technologies mostly, and now I desire to spend more efforts on learning how to harvest the data in order to get the best business values out of it. My expectation from this masters program is to help me to enhance my capabilities in the areas of data analytics, predictive modeling & statistics which I currently do not own significant experience.
Dr Gert Janssenswillen shared information about his research on business engineering in Hasselt University in Belgium in the title of “Enhancing discovered process models using Bayesian inference and MCMC”(Markov chain Monte Carlo methods). He explained Process Mining as process oriented data science; such as a process happens somewhere in the company generate event log activities and from that data discover process models are produced to see which algoritm works which kind of data. He gave examples to make the audience see how complex the traces can be and how hard it is to be explained by their bayesian model when the probablities and dependencies comes in. They use BupaR the central package which consist 8 packages for handling and analysis of business process data. It supports the different stages of process analysis such as importing event data, calculating descriptives, process monitoring and process visualization.Propro, the probabilistic process package takes the process data and the process model and gives the bayesian model to see u can do certain hypothesis or not.
In the app world for marketing location data analysis is really significant, that’s why the article got my interest. The author shows a few ways to visualize the location history collected and provided by Google history with R. The article shows how to get the distribution of the data over periods of time, get the accurancy of the data plus plotting the data points on maps.
#2- How to cluster your customer data — with R code examplesThe article gives information about k-means clustering and Agglomerative Hierarchical Clustering use on the customer sales data both in theory and in practice. The R codes for the algoritms are shared on github page. Customer segmentation is a key point for the offering management in marketing.
#3- Understanding and Writing your first Text Mining Script with RText mining reminds itself to me when I am busy with the product data which comes from different partners. In this article the author tells basicly how to load, clean, stem the data and create term matrix using R.