exorbitant data & intergalactic machines


exorbitant data & intergalactic machines


“Data! Data! Data! I can’t make bricks without clay!”

– Sherlock Holmes –

My name is Jochen Roeth (M.Sc. Data Science) and “Exorbitant Data and Intergalactic Machines” is my project portfolio about Data Science, Deep Learning, AI,  Big Data Technology and Future Trends. This includes Data Exploration and Data Visualization, Statistical and Machine Learning. Furthermore about Data Protection and Data Privacy, ethical aspects in the field of Big Data and Algorithms. Essential for these things is a Business Understanding of the data and new concepts of Business Models such as Platform Ecosystems, Sharing Economy and Smart Data Products. If you got any questions, hints or recommendations, please don’t hesitate to drop a comment.

project portfolio

Text Mining in Finance

Building a sentiment index in R to predict disruptions in financial markets with Twitter API

Recommender System

Building a digital platform ecosystem “skills engine” to connect people with right skills to get swift technical solutions. Natural Language Processing with Word Embeddings (Word2Vec). Cooperation with University of Applied Science Darmstadt @ Devoteam GmbH

Preventive Maintenance

Building a Hierarchical Temporal Memory (HTM) unsupervised real-time anomaly detection algorithm for streaming data for preventive maintenance @ Robert Bosch GmbH

Big Data Batch- and Stream Processing

Processing and analysis of large batch and streaming data on a distributed computing cluster with Hadoop/MapReduce, Apache Spark and MongoDB

Computer Vision

Building a object detection and localization algorithm for image and video data with Convolutional Neural Networks (CNN) and YOLOv2 (car detection) with Python, TensorFlow and Keras

Preventive Maintenance

Building a semi-supervised Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) Cells and bayesian hyperparameter optimization for anomaly detection in time-series sensor data @ Robert Bosch GmbH with Python and TensorFlow

Trigger Word Detection

Natural Language Recognition Trigger Word Detection with Long Short-Term Memory (LSTM) with Python and TensorFlow

AI business strategy planning

AI business strategies planning and business model optimization for AI engineers and porjects

Computational Social Science

Model and simulate individual or collective entities with Agent Based Modeling (ABM) to detect emergent phenomena in social systems with NetLogo. Moreover graph analysis, text mining and survey science.



Darmstadt, Germany, Earth, Milky Way
+49 (0)177 8415541