DHKT

External Seminar Of FSI: “Machine Learning” - Professor Rodica Lung from Babes-Bolyai University

27/04/2023

Within the framework of the Scholar Exchange Program with a partner university in the Erasmus+ project, on the morning of April 26, FSI students had an exchange with Prof. Rodica Lung, Faculty of Statistics, Forecasting and Mathematics, Babes-Bolyai University, Romania with the topic “Scientific research methods and applied Machine Learning Algorithms”

During the seminar, Professor Rodica Lung extended her warm greetings to the students of the Faculty of Statistics and Informatics. She expressed her great delight at the opportunity to meet and engage in discussions with the students during her lecture.
According to GS. Rodica Lung, the field of machine learning is interested and talked about by many people, therefore, there will be many success stories related to machine learning.
She presented her point of views on how well an algorithm can learn from enormous volumes of data in a particular domain, how accurate the algorithm is with errors in its modeling assumptions or errors in the data training the ability to design an efficient algorithm based on a given problem. Furthermore, GS. Rodica Lung believes that privacy protection derived from data processed by ML algorithms should not be underestimated.
In addition, to create truly meaningful and widely acceptable systems, it is equally important to take into account the interaction between humans and AI, correctness and ethics in application.


Professor. Rodica Lung talks about Machine Learning

According to GS. Rodica Lung, best practices for evaluating machine learning models based on those 2 points: Statistical validation results and how to classify the dataset as training data, test data.

Finally, GS. Rodica Lung offered a few lessons to make data science better: Although we agree that the popular methods in applied machine learning research have produced many highly successful products, that doesn't mean we shouldn't keep striving to do better. More stringent norms should be introduced for statistical analysis, testing and recording of datasets during training, and comparison of results with other models. In addition, she encourages the development of necessary norms and attitudes by asking these questions during the peer review process, questioning them to the project leader and making your personal commitment to them. Even if people aren't interested in your question, raising and normalizing standards issues as a point of discussion is an important first step.

Students are enthusiastic during the lecture and have numerous questions.

After the presentation of Prof. Rodica Lung, the students also asked many questions about new methods to improve data reliability and the importance of machine learning research in the community. Students are very excited to discover a new field, which will bring many benefits in learning as well as better understanding of future career opportunities.

.

Dr. Hoang Thi Thanh Ha gave a gift and thanked to Prof. Lung. 

Dr. Hoang Thi Thanh Ha - Vice Dean of the FSI thanked Professor Rodica Lung, and hoped that, in the coming time, Prof. Rodica Lung will spend more time to exchange and present many useful lectures for students of the FSI in particular and students at the University of Economics in general.


Source: Center for ICT