Christian A. Hammerschmidt

I am a PostDoc in the Cyber Security Group at the Intelligent Systems Department at the TU Delft. I work on machine learning methods for reverse engineering. Together with  Sicco Verwer, I apply these methods both in network security and software testing.

Together with Sicco Verwer, I am also working on APTA Technologies, a spin-off company offering services and solutions for analyzing sequential, computer-generated data.

Before coming to Delft, I successfully defended my Ph.D. thesis at the Interdisciplinary Centre for Security, Reliability and Trust, which is part of the University of Luxembourg. I spent my time in the Service and Data Management (SEDAN) group headed by Radu State. I still regularly work with SEDAN.

About my Work

I am interested in machine learning and its applications, in particular in protocol and program inference, and program analysis with applications in software engineering and cyber-security. More specifically, and I want to learn a model that converts a sequence of observations (or inputs) into a sequence of responses (or outputs). Ideally, the model is not just a magic black box, but is a description of an operational or computational process, explaining how to convert the input to the output. With that in mind, I like to think about interpretability and explainability. Recent work of interest in the field of machine learning includes neural models for program synthesis (and to a lesser extent, also induction).
Recently, I am also working, though to a lesser extent, on learning representations of discrete and structured data.

During my Ph.D. in Luxembourg, I focused on learning models that at their core have a finite state space: (probabilistic) finite state machines, deterministic transducers like Mealy machines, and ad-hoc defined variants of finite models. Despite their limited expressive power (compared to Turing complete universal computation models), there are many useful applications of these models in computer sciences. Together with Nino Pellegrino and Qin Lin at TU Delft, we looked at network traffic and summary statistics to build models that profile and fingerprint hosts on networks.
I also built on Sicco Verwer’s DFASAT tool to build flexfringe, a tool to define heuristics to learn a wide range of automaton models using a red-blue type of state-merging algorithm.

Recent Activities

Together with a team of colleages from TU Delft, we won the KDD Adversarial Machine Learning Workshop’s Robust Learning challenge both on the attack and defend tracks. Details of the results are available online.


Recent publications:

See my ORCIDGoogle Scholar, and pages.

Besides academic work and publications, I am also trying to blog over at and my personal blot, cinrizasti.


Talks given in 2019

  • Generative Neural Models for discrete and tabular data at the Cyber Security seminar at TU Delft

About my Background

After passing the undergrad portion of my computer science diploma degree at the FAU Erlangen-Nürnberg In Germany, I concentrated on mathematical methods and theoretic foundations in computer sciences, taking classes on numerics of partial differential equations, functional analysis, simulation on high performance systems as well as algebra, information and quantum information theory, logic, complexity theory, and cryptography.

Towards the end of my program, I moved to Berlin and joined Uwe Nestmann’s group on models and theory of distributed systems as a student researcher. Both my undergrad and graduate (“Studienarbeit” and “Diplomarbeit”) theses were written during my time in Berlin, where I worked on a project investigating the expressive power of distributed modeling languages, in particular variants of Petri nets, CCS, pi calculi, and the join calculus. In my theses, I looked at reactive systems and contextual transition systems of Robin Milner’s bigraphs. I encoded some calculi into bigraphs in such that the encoding respects not only bisimulations but accounts for also step-readiness semantics.


I enjoy a solid balance between working in the abstract, i.e. using formal systems and investigating their potential and limits, and having impactful, hands-on results in the real world. Answering questions about the expressive power and capability to predict or explain the world fascinates me as much as having a practical way to exploit the answers to these questions.

The draw of the abstract and mathematical, for me personally, lies in the flexibility of their applications, which range from the social sciences to engineering, from bio-informatics to cognitive science, from physics to finance. My curiosity and passion for applying mathematics in a transdisciplinary way have led me to pursue psychology, linguistics, and politics.

In my free time, I enjoy learning new languages (currently, French and Dutch, you can check out my profile on Duolingo), reading (especially science fiction, thrillers and fantasy novels), and athletics (I particularly enjoy skiing whenever I have the chance).