Can Berk profile photo

Can Berk

Machine Learning Researcher | Data Analytics
Oldenburg, Germany can94berk@gmail.com +49 1789771249 LinkedInGitHub
Data analytics graduate specialized in machine learning, data analysis, and computational modeling. Experienced in developing predictive models and working with structured and unstructured data, with research exposure to neural networks, signal representations, and speech synthesis.
Experience

Master Thesis: Similarity-Infused Graph Neural Networks for Time Series Classification

Hildesheim University
Thesis grade: 1.0
  • Proposed a graph-based time series classification model using GNNs, representing time series as nodes connected by similarity-based relationships, achieving competitive performance with state-of-the-art methods on benchmark datasets.

Machine Learning Tutor

Hildesheim University
Hildesheim, Germany
  • Graded assignments and exams, providing constructive feedback to students.
  • Presented detailed solutions for weekly machine learning assignments, including coding exercises and theoretical concepts.

Data Science - Workstudent

Statista
Hamburg, Germany
  • Deployed a natural language processing (NLP)-based company entity matching model, increasing accuracy by 25% compared to the previous method.
  • Implemented data pipelines and improved data quality using Python.

Research Project: Speech Synthesis and Translation via Exemplar Autoencoders

Hildesheim, Türkiye
  • Implemented and experimented with autoencoder architectures to explore latent representations of speech signals for speech synthesis and translation tasks.
Education

MS in Data Analytics

Hildesheim University
Hildesheim, Germany • GPA 1.4
  • Coursework on statistical and mathematical models in data analytics, computational methods and data programming in Python.

BA in Economics and Business Administration (Double Major)

Koc University
Istanbul, Türkiye • GPA 1.3
  • Coursework on Econometrics, Statistics, Quantitative Methods in Business, Management Information Systems.