COFFEECONTALK S1 Ep 10
What is COFFEEECONTALK?
COFFEECONTALK is an interview series which is posted on GIPE’s undergrad YouTube channel on the last day of every month. It is a part of The 8:10 newsletter’s Econcordia section and thus revolves around Economics, Business and Finance.
Here’s what’s in store for this episode!
For this episode, we had the pleasure of inviting Mr Dev Kakde . Mr. Kakde is a statistician with education and expertise in Manufacturing and Transportation. He has also contributed to original research in machine learning theory, quality control and prognostics. He is also an ASQ certified reliability engineer and has a knack for solving engineering problems using statistics and machine learning techniques. Currently, he is working as a principal research statistician and developer with SAS analytics.
In this interview, Padmaja and Wani have a very insightful dialogue with Mr Kakde on the momentousness of data and statistics. Mr Kakde also shares his tips on how economics students interested in data science can pursue this field.
Following are the questions and their timestamps:
- (0.42) Wani: Nowadays, we hear a lot about how DATA IS THE NEW OIL, how valuable yet useless it is in its unprocessed form. With your background in manufacturing and transportation, we would love to hear your insights on the importance of data analytics in this sector.
- (7.22) Padmaja: You have a knack for solving engineering problems using statistical and ML techniques. Would you like to tell us about some of your projects?
- (13.26) Wani: AI is increasingly being used in making business decisions. To what extent do you think it can replace human efforts?
- (18.48): Padmaja: Recently the FTC has issued a report warning saying that AI solutions can be inaccurate, biased, and discriminatory by design and hence shouldn’t be relied on as a policy solution. What is your opinion about it? There have also been some controversies like the one involving Microsoft. Their AI chatbot, Tay, which was an experiment in conversational understanding, was coming up with its own version of hateful speech after absorbing people’s conversations. Now, machines certainly don’t think for themselves, they simply process what input they get to come up with their output. So who is responsible for/ how can we regulate what input goes into these machines? Has any progress been made on that front?
- (22.46) Wani: There is a belief that statistics focuses a lot on the behavior of the data model rather than the data and processes that may have generated it. In short, it focuses on making the data fit the model rather than adapting the model to fit the data which is where the algorithmic modeling culture of machine learning can come in handy. What are your opinions on this and the relationship between statistics and machine learning?
- (25.46) Padmaja: Last but not the least, we have noticed a trend even with our batchmates, that the economics graduates pursue a degree in data sciences sometimes even side by side with graduation . So, I wanted to ask, according to you, what are the future prospects for the students with an economics background and how should one decide if they want to pursue a combination of economics and data sciences?