Florian Muellerklein

Data Science


Neural Networks
Deep Learning for Computer Vision
Deep Learning for Natural Language Processing
Teaching and Presentations



B.S. Psychology, UMBC



Recent Projects

Social Media Analysis Company
Identifying Named Location Entity Information from Text
Implemented a deep learning based named entity recognition model using a bi-LSTM with a conditional random field. This model is used to determine ambiguous references to locations in natural language human text in multiple languages.

Government Railroad System
Crack Detection and Defect Detection in Jointbars
Utilized RFCN, Densenet, and Faster-RCNN deep learning approaches to identify cracks, missing bolts, and missing nuts from black and white images of joint bars. Developed a system for SMEs to annotate images quickly by suggesting initial bounding boxes using the deep learning models.

Technical Expertise

  • Convolutional Neural Networks
  • Recurrent Neural Networks
  • Machine Learning
  • Deep Learning Frameworks: Tensorflow, PyTorch, Keras, Caffe
  • Python data stack: scikit-learn, pandas, numpy, Jupyter, others
  • Apache Spark and PySpark