Introduction to myself

My name is Barış Sivas and I am from İzmir originally. I live in İstanbul since 2011 after coming to the city to go educational career with my college degree.

I graduated from Istanbul Technical University Industrial Engineering department in June 2017. As a fresh graduated, I started to work for Turkcell Company as product manager of fixed enterprise data products. After spending 2.5 amazing years, I decided to change my role in the same market with different perspective. Therefore, I accept a job offer from Innova, which is Sub Company of Turk Telekom as Business development specialist.

After I started working for Innova, I had the opportunity to take a closer look at Artificial Intelligence and big data management projects, which is one of the fields of study. Both the content and the target results of these projects affected me so much that I decided to start research on this field. I aim to shape my career journey by combining my knowledge that I have gained in sales & marketing with a different perspective as a post analytics master of analytics.

You can see my linkedin profile from here

Use R!2020 Virtual Conference: An R Package for Interpretable Decision Tree Visualizations (Trang Le)

In this video “An R package for interpretable decision tree visualizations” is presented by Le Trang as postdoctoral presentation at University of Pennsylvania. She suggests that you can transform your decision trees into a cool and hot visualization with using heat map.

At first, she introduces some kind of R packages like Rpart.plot, visNetwork::visTree(), Partykit::plot.party(), Ggparty: (histogram / density plot) and then she continues her speech to talk about package of treeheatr and which is kind of decision tree model that is predicting different species of Covid-19 impacts. The main idea of this presentation is about the effect of heat map to decision tree visualization in addition to the R packages. In my opinion, the part of video when the examples are given is very helpful to understand for efficiency is this.

You can see the presentation from here

Three R Posts Relevant To My Interest

R Can Pull The Fire Alarm!

This article is written about a topic that is it possible to manage your daily life in fully automated and to be notified by R codes via e-mail, text message, slack message or Microsoft Teams message. You can see the details of R codes in the post from here

Monitor COVID-19 at the COVID-19 Forecast Hub

Nowadays it is an undeniable fact that every person in the world worries about the number of effected people from Covid-19 diseases. Therefore, the war against these virus crises is getting larger day by day in different areas. According to Joseph Rickert who is the writer of this article, The Reich Lab Covid-19 Forecast Hub (University of Massachusetts) is one of the best option to look and monitor expert forecasts for United States weekly and cumulative COVID-19 deaths. Besides, every Tuesdays, four-week, national and state level forecasts are published via Hub from over thirty-five different groups along with its own ensemble forecast. In addition, these forecasts are available in GitHub Repository for interactive visualization.

You can see the details from here

Modern Rule Based Models

According to the some scientists, we are currently in a new era, which is called as Digital Era. It is explained with various causes however, the most exciting one is Deep Learning. The technology is able to develop itself thanks to its special power, which is known as not only “understanding of data” but also “producing decision models”. Of course, we cannot forget the role of machine learning models especially in this phase. There are many different shapes, sizes and classes of machine learning models that effects to various problem domains. This article is written by Max Kuhna to give a summary of several rule-based models such as C5.0, Cubist and Rulefit.

You can see the details from here