At 10:00 am, June 14, 2023, Professor Guotai Chi, School of Economics and Management, Dalian University of Technology, was invited to give an offline lecture entitled "Early Warning of Small Enterprise Default Risk: Deep Learning of Index numerical Transformation". Professor Minggui Yu, Dean of the School of Finance and co-director of the Digital Technology and Modern Finance Innovation and Intelligence Base of Zhongnan University of Economics and Law, presided over the lecture. All graduate students of the School of Finance and researchers of the base attended the lecture.At the beginning of the lecture, Professor Minggui Yu briefly introduced Professor Guotai Chii's relevant information to the participants, and expressed his sincere thanks for his ability to conduct lectures for our institute in his busy schedule.

This lecture was introduced by Professor Guotai Chi from five aspects: research background, significance and innovation, research status, default discrimination model construction, empirical analysis and research conclusions. First of all, he summarized the significance of small business credit risk identification, introduced two scientific propositions and the similarities and differences between propositions and "blueprints". Further, the innovation points and research findings are introduced, and the method used in this paper is significantly improved compared with the traditional method.

Professor Guotai Chi proposed that the samples used in the construction of default discrimination model are mainly divided into training samples, verification samples, test samples and balanced training samples. The innovation of this model mainly focuses on the optimal index combination selection based on linear support vector machine and the index numerical transformation based on deep neural network. Then, Professor Chi Guotai explained that this study is better than deep learning model, machine learning model and statistical model according to the actual results

Finally, the participants had an active academic exchange with Professor Guotai Chi, who gave detailed answers to the questions and perplexities of the teachers and students. After the exchange, Professor Minggui Yu expressed his sincere thanks to Professor Xu Wei for his wonderful sharing. I believe that the teachers and students attending the seminar must have a deeper understanding of deep learning and small business default prediction analysis. I hope students can apply what they have learned to the subsequent scientific research and expand the research ideas and directions.