Hvis du har foretaget ændringer i Pure, vil de snart blive vist her.

Personlig profil

Forskningsinformation

Christian M. Dahl's research agenda is broad and interdisciplinary. One part is focusing on theoretical statistics and time series analysis with applications in finance. A second part is focusing on nonparametric estimation and causal inference, including random fields, neural networks and deep learning.  Christian has extensive experience working with deep neural networks and random fields (computer vision and machine learning techniques) and has made significant theoretical and empirical contributions to this research field. With the increasing availability of an exploding number of digitized historical documents constituting an incredible new, rich and for large parts still unused source of data Christian is directing his research and his expertise in computer vision and deep leaning towards developing tools for automated harvesting (transcribing) data from digitized historical (and possibly degraded) documents. Christian has established an interdisciplinary research unit on “Big Data Analytics and Digitization” (BDAD) within the Department of Business and Economics (DBE) at SDU. The aim of this research unit is to become a leading interdisciplinary research unit worldwide in harvesting, linking and analyzing individual data from digitized/scanned structured handwritten historical documents. Current research projects involve developing deep learning techniques for transcribing census and church records and about 20 million Danish death certificates. Christian and members of BDAD are currently working jointly with many national archives in developing citizen science data collection strategies (see, for example, www.basecrowd.dk) to begin digitalization of health data, including the Danish death certificates.  The harvested data will make it possible to study early-life health policies, parental investments, the formation of human capital, and socio-economic and health trajectories. CMD has also recently received funding from the InnoBooster program regarding the development a web-based software application that based on AI and Machine Learning tools can assist citizens in evaluating (probabilistically) their prospect in case of lawsuits.

Forskningsområde

  • Big Data Predictive Analytics
  • AI and Machine Learning
  • Computer Vision
  • Time series Econometrics
  • Quantile Regression
  • Panel Data Analysis

CV

Please consult the english CV

Fingeraftryk

Fingeraftryk er baseret på at anvende teksten fra personernes videnskabelige dokumenter til at skabe et indeks med vægtede emneordskoncepter, som definerer de vigtigste emner for hver enkelt forsker.
  • 4 Lignende profiler

Netværk

Nyligt eksternt samarbejde på landeniveau. Dyk ned i detaljerne ved at klikke på prikkerne.
Hvis du har foretaget ændringer i Pure, vil de snart blive vist her.