Python
Andrés Cabero Busto
Working Student
Stuttgart
Overview
1
1
year of professional experience
7
7
years of post-secondary education
4
4
Languages
Work History
Student Researcher
University of Stuttgart
Stuttgart
08.2022 - Current
- Development of a benchmark tool for the field of Lexical Semantic Change Detection (https://github.com/garrafao/LSCDBenchmark)
- Daily use of contextualized word embedding (BERT, XLM-RoBERTa, RoBERTa) via the HuggingFace Transformers library.
- Tool based on Hydra (https://hydra.cc/) for easy experiment setup and configuration.
Working Student
Sony
Stuttgart
07.2022 - Current
- Development of a natural language grammar for data normalization
- Data annotation of audio samples
Education
Bachelor of Arts - Linguistics & NLP
09.2016 - 12.2020
Got introduced to Computational Linguistics - Natural Language Processing in my second year of my bachelor´s degree
Bachelor of Arts - Linguistics & NLP
09.2019 - 07.2020
Became an exchange student in Madrid for one year, because the Linguistics program there has several more subjects related to NLP and Computer Science. Grades: 9.9/10 in programming for NLP, 9/10 in a general NLP subject.
Master of Science - Natural Language Processing
04.2021 - Current
Skills
Time Management
Software
Rust
TypeScript
Go
Timeline
Student Researcher - University of Stuttgart
08.2022 - Current
Working Student - Sony
07.2022 - Current
University of Stuttgart - Master of Science, Natural Language Processing
04.2021 - Current
Universidad Complutense De Madrid - Bachelor of Arts, Linguistics & NLP
09.2019 - 07.2020
University of Cádiz - Bachelor of Arts, Linguistics & NLP
09.2016 - 12.2020
University Projects
Explainable AI: Application of XAI methods to an LSTM network trained to predict emotion categories in speech samples using the Captum library. Of the employed methods, SHAP produced the best results, as we were able to retrain the network on a small subset of features deemed relevant by SHAP, while retaining the performance of the full model. In the end, the number of parameters of the model was cut by more than half.
University Projects
Explainable AI: Application of XAI methods to an LSTM network trained to predict emotion categories in speech samples using the Captum library. Of the employed methods, SHAP produced the best results, as we were able to retrain the network on a small subset of features deemed relevant by SHAP, while retaining the performance of the full model. In the end, the number of parameters of the model was cut by more than half.
Similar Profiles
Swetaki ChatterjeeSwetaki Chatterjee
Research Associate at University of StuttgartResearch Associate at University of Stuttgart
Mohammadhassan SamadpourMohammadhassan Samadpour
Research Associate at University of StuttgartResearch Associate at University of Stuttgart
Emilie IsmailovaEmilie Ismailova
Research Assistant at University of StuttgartResearch Assistant at University of Stuttgart
Samruddhi SatheSamruddhi Sathe
HiWi (Student Assistant) at University of StuttgartHiWi (Student Assistant) at University of Stuttgart
Angelique GonzalezAngelique Gonzalez
Childcare Assistant at Stuttgart Gospel ServiceChildcare Assistant at Stuttgart Gospel Service