Tetiana Kodliuk is a Data Scientist at V.I.Tech in Lviv. After completing her Ph.D. in Mathematics, she worked as a lecturer at the Ternopil National Pedagogical University and as a lecturer of a Data Analysis online course at the Mumbai Aegis School of Business. In the interview for IT ID Lviv magazine, Tetiana explains how she still uses math daily after 8 years of experience.
It all started with math. It was my favorite subject at school. I loved math so much that I completed every math assignment for the whole year in September already. Later, at university, I realized I wanted to use math in real life, so in my 4th year, I started to work as an analyst at a local company. It was very interesting for me to do data analysis because there was a lot of information that needed to be understood, visualized, structuralized, and processed.
After graduation, I decided to continue to study math, so I started a Ph.D. at the National Academy of Sciences of Ukraine in Kyiv. It was very interesting, but it was still very theoretical. My thesis was about differential equations. In my research, I tried to find an applicable general equation which could describe physical processes, but I didn’t end up using it in real life. My work was very exciting, but I kept asking myself the same question: how can I apply this theory to reality?
Data Science as a combination of math, analytics and IT
People tend to joke that women start working in Data Science while on maternity leave. There is some truth in this because when you work, you don’t have free time to learn new things. In my case, the maternity leave gave me the time to learn. I completed around 20 online courses. The main courses were Machine Learning by Andrew Ng, Data Science on Coursera, Big Data and Apache Spark on edX. Unfortunately, Ukrainian universities don’t offer these specific courses, so doing them online was the only option.
The more learn, the more I understand there’s still so much I don’t know. It’s an infinite process, but I like it. Being a Data Scientist makes it possible to self-develop constantly. Without that continuous self-growth, I would be bored.
At V.I.Tech, we are developing a platform for American hospitals based on Data Science. At the moment we have around 54 million users in the US. We analyze personal data to predict any future illness for a specific patient. For example, we can predict what will happen with a specific person’s health on the 16th of July in 2020. It’s incredible! This data includes demographic information, nationality, as well as the doctor’s notes from the patient’s visits, etc. Big Data, together with Machine Learning and Deep Learning, can analyze a huge amount of information. Based on this information, a computer divides all patients into groups with similar criteria. Comparing this data allows us to predict the future of a person.
Obviously, health is very important, so the accuracy of the prediction should not be less than 90%.
The final decision can’t be made by a computer – it’s a basic rule in medicine. The computer’s research serves a recommendation. The human mind, working together with AI, is more efficient, however, only a person can make the decision. This data is not available to the public. All 70 people working on this project at V.I.Tech completed their training and received HIPPA certificates in information confidentiality.
We can predict almost everything
Data Science makes our lives easier. We use for work optimization, time management, and the automation of routine jobs. Data is useful in every aspect of our lives. Nowadays, Data Science is mostly used in sales and services, energy and automotive industries, education, and medicine. Financial markets and media also benefit from it. The amount of information is increasing at such an extreme rates that the human mind can’t keep up with it, but an algorithm can.
Analyzing social network data can predict what a specific person will do at a certain place and time, what a person will decide to buy, when they will want to quit their job or when they will be sick.
Sometimes I have the feeling that my job really matters. Especially if we are talking about medical research. It is extremely rewarding to I know that my job is really needed. I’m basically saving people. The moment this platform will be launched, it will really help a lot of people.
Speaking at conferences makes me feel like a teacher, and I really like it. It’s very interesting to see what people are asking – it gives me the feeling that I’m not doing my job for anything. It’s not enough for me to work on a project for the whole year and only communicate about it with the client. At conferences, I have the opportunity to share my knowledge and get some feedback. It inspires and motivates me to continue my work. On top of that, it’s the perfect place to expand my knowledge too. The expertise of other speakers usually inspires me to try new technologies and methods.
In collaboration with my colleagues, I created a Data Science course for the Mumbai Aegis School of Business. I’ve made online lectures for students and evaluated their assignments. Even though most of the students are already working, they still occasionally write me an email. I will meet some of them at the Data Science conference in Mumbai soon.
The future of Data Science in Ukraine
I want to do something important here in Ukraine. I’m very optimistic about the future of my country. Most of the IT and Data Science specialists are working for outsourcing companies, but they are still living here and some of them are planning to launch their own projects and products. We already have a few Ukrainian startups that went global.
In Ukraine, Data Science is already applied in some projects in medicine. Recently, EHealth was launched – a project which aims to computerize data in hospitals. Lviv, Kyiv, and Kharkiv have already implemented projects that analyze ambulance data.
Lang-uk is another promising project in Ukraine. With the help of Machine Learning, a group of enthusiasts is working on a program which will identify the Ukrainian language. Together with Dmytro Chaplynskyi, the project leader, we’ve made a model which transfers every Ukrainian word to an n-dimensional vector and uses it for the classification of texts in Ukrainian, as well as sentiment analysis, clustering, entities recognition, etc. The results of our research are publicly available.
Success comes to those who work hard and put their soul into their job. Everyone in Ukraine can accomplish their own goals, you don’t have to wait for someone to teach you, you can do it yourself. Self-development is the main key to success.