Assignment 1 – Task 3

Video Summary & Data Analysis

Industrial Engineering and Data Science integration based on Cem Vardar’s interview.

Summary

I watched the interview with Cem Vardar, who holds a PhD from Arizona State University and has worked at companies such as Intel and Carvana.

The interview mainly focuses on the relationship between Industrial Engineering and Data Science in real-world applications.

According to Vardar, Industrial Engineering is fundamentally about solving problems within complex systems, while Data Science should be viewed as a practical tool for business problem-solving rather than a standalone profession.

One of the most important concepts mentioned in the interview is “Evolutionary Design,” which emphasizes improving systems gradually through small, data-driven changes instead of replacing entire systems at once.

He also highlights that foundational tools such as SQL and Excel remain highly valuable despite rapid technological changes.

Questions

  1. Open-Ended Question

What is Cem Vardar’s “Evolutionary Design” approach, and why does he suggest it instead of replacing an entire system?


Answer

It refers to improving systems incrementally through small and manageable changes. This approach is preferred because complete system replacements are often expensive, risky, and difficult to implement successfully in real business environments.

  1. Multiple-Choice Question

Which tool is described as a “must-have” skill for over 20 years?

  1. Neural Networks
  2. Deep Fakes
  3. SQL (Structured Query Language)
  4. ChatGPT

Answer

  1. SQL

AI Disclosure

I used Gemini 3 Flash to:

  • Summarize the interview transcript
  • Format the questions
  • Improve clarity and readability

I also used Gemini AI to debug and correct errors in my R code.

Video Link

Watch the Interview on YouTube

Data Analysis Section

In this section, the polls_us_election_2016 dataset from the dslabs package is analyzed using R.

library(dslabs) library(dplyr)

data(“polls_us_election_2016”) data <- polls_us_election_2016

my_first <- “Bora” my_birth_year <- 2004

k <- (nchar(my_first) + my_birth_year) %% 15 + 8 k

if (k %% 2 == 0) { head(data, k) } else { tail(data, k) }

sum(is.na(data))

na_per_column <- sort(colSums(is.na(data)), decreasing = TRUE) head(na_per_column, 8)

new_data <- data %>% mutate( across(where(is.numeric), ~ ifelse(is.na(.), my_birth_year + k, .)), across(where(is.character), ~ ifelse(is.na(.), paste0(my_first, “”, k), .)), across(where(is.factor), ~ { x <- as.character(.) x[is.na(x)] <- paste0(my_first, ””, k) x }) )

if (k %% 2 == 0) { head(new_data, k) } else { tail(new_data, k) }

sum(is.na(new_data)) anyNA(new_data)

Built with R and Quarto

```

Back to top