I am Data Scientist with a background in Econometrics and Operations Research. Specialized in Predictive Modeling, Causal Inference, and Optimization methods. Professional experience in all Machine Learning sub-disciplines, including NLP, Computer Vision, and Recommender Systems. Committed to consistency and transparency, dedicated to continuous improvements, and excited for new challenges.

An object detection model to predict person in indoor and hospital settings with a custom synthetically created dataset in Blender. An innovative solution to detect person instances without requiring actual person images by utilizing the state-of-the-art Bayesian CycleGAN to tackle down the challenging Synthetic-to-Real translation task.
Language: Python, Blender
Date: Q1 2022

Extending the state-of-the-art Time-Series forecasting models by utilizing a data-driven anomaly detection and complex seasonality clustering approach using Self-Organizing Map and Hierarchical Clustering methods. Results showed significant improvements on the existing forecasting models based on a 1-year ahead prediction of multiple time-series of over 10 TV-channels.
Language: Python
Date: Q1 2021

Developed an advanced Gaussian Mixture Model from scratch by generating samples from a mixture of Gaussian distribution, improving the existing Machine Learning clustering methods and allowing for more complex clustering.
Language: Python
Date: Q4 2020

Created a general framework of how to construct hybrid models making full use of linear and non-linear models applied on a multiple time-series dataset. Key concepts such as outlier detection, correlation matrices and imputing missing datapoints are also discussed here.
Language: R
Date: Q3 2020

Build a Recommender system based on Collaborative Filtering and extending it with a Facotrization Machine incl. ALS, RW. Suggested an alternative evaluation metric on measuring the performance of Recommender Systems in general.
Language: R
Date: Q2 2019