Oğuz Ay

I have been working in Garanti BBVA Teknoloji as a business analyst for 8 years now. I am working in investment banking, so the main products we work fora re investment products like equities, mutual funds, treasury bonds, etc.

There are many times we encountered and gave meaning to customers’ investent data. That’s when i understood the importance of data and wanted to focus more in data science for my future career.

UseR-2019 Video Review

Link: https://www.youtube.com/watch?v=DlppjRYVklQ

Jesse Islam made a presentation on Biostatistics & Epidemology mainly focusing on forecasting on prostate cancer.

At the beginning of the presentation, he mentioned about Cox Regression models which is one of the most popular methods in survival data analysis. He used European Randomized Study of Prostate Cancer Screening data which consists of almost 150,000 men that are 55 – 69 year of age. He used casebase method for determining the risk of prostate cancer using logistics regression.

Apart from Cox, he used Exponential, Gompertz and Weibull distributions for result comparison. As a result, he found out that all four methods gave similar standart errors and results. He then calculated the Cumulative Incidence (Absolute Risk) in order to conclude the cancer risk. As the absolute risk function showed, he concluded that the individual he examined did not have a risk for upcoming 2 years.

R Related Posts

1. Link: https://towardsdatascience.com/analyzing-stocks-using-r-550be7f5f20d

Summary: In this article the author, Joy Gracia Harjanto, searched for Amazon’s stock (AMZN) which increased by 95.6% in 2017. She started her analysis by first visualizing the price of the stock itself. She then used Bollinger Band chart, % Bollinger change, Volume Traded and Moving Average Diverence in 2018. She then compared the price of Amazon in comparison to other technology firms such as Facebook, Google and Apple. She then made predictions on the price of Amazon based on random walk theory and monte carlo method.

2. Link: https://analyticsprofile.com/algo-trading/cryptocurrency-data-analysis-using-r/

Summary: In this article, the author (Sai Swapna Gollapudi) compared and analysed the crypto currencies with each other. She looked at the market cap, price change, transaction volüme, market cap trend, price movement and market cap growth of these crypto currencies

3. Link: https://rstudio-pubs-static.s3.amazonaws.com/239816_eab2216881594f939a7e0e520979387a.html

Summary: In this article, the author (Bolor Boldbaatar) mentioned about a recommendation tool for bank products. She first cleaned her data to use in analysis. Then she made some exploratory analysis (histogram of the customers age, age distribution over years, the customers’ location in map view) She then looked at the income of the customers according to their segmentation. She shows the customers’ correlation to the various bank products. She then used data in Recommenderlab package for User-Based Collaborative Filtering for prediction model.