Klaus Ackermann

Klaus Ackermann is a Senior Lecturer (Assistant Professor) in the Department of Econometrics and Business Statistics at Monash University in Melbourne, Australia. His research interests are in the areas of Artifical Intelligence, Causal Inference, Applied Econometrics, and Alternative Data.

He holds a PhD in Economics from Monash University and BSc and MSc in Business Informatics with major in Economics from the Technical University of Vienna. He pursued a postdoctoral fellowship at the Center for Data Science and Public Policy at the University of Chicago.

Klaus is a founding member of Monash SoDa Labs, an empirical research laboratory associated with Monash University’s Department of Economics and Department of Econometrics in the Monash Business School. SoDa Labs applies new tools from data science, machine learning, and beyond to answer social science questions using alternative and big data.

Klaus is also the co-founder and one of the directors of KASPR Datahaus Pty. Ltd. and co-founder of the IP Observatory.

Publications

Causal Inference and Machine Learning methods

Grecov, P., Prasanna, A.N., Ackermann, K., Campbell, S., Scott, D., Lubman, D.I. and Bergmeir, C., 2022. Probabilistic Causal Effect Estimation With Global Neural Network Forecasting Models.  IEEE Transactions on Neural Networks and Learning Systems. [Published Version] 

• We propose an estimation procedure for estimating distributional treatment effects in a synthetic control style setting. Rather than estimating the average effect only, we estimate the full distributional treatment effect for causal inference in the distributional tails.

Grecov, P., Bandara, K., Bergmeir, C., Ackermann, K., Campbell, S., Scott, D. and Lubman, D., 2021, May. Causal Inference Using Global Forecasting Models for Counterfactual Prediction.  In Pacific-Asia Conference on Knowledge Discovery and Data Mining (pp. 282-294). Springer, Cham. [Published Version] 

• We propose a new causal inference method similar to the synthetic control methods. Instead of estimation of a weight matrix, we make use of time-series properties and forecast the average treatment effect.

Economics

Ackermann, K., Churchill, S.A. and Smyth, R., 2024. Estimating the relationship between ethnic inequality, conflict and voter turnout in Africa using geocoded data. World Development, 180, p.106644. [Published Version] 

• We examine the relationship between ethnic inequality, conflict, and voter turnout in Africa, finding that higher ethnic inequality increases conflict, which in turn reduces voter turnout, with this effect being more pronounced in highly populated areas.

Ackermann, K., Churchill, S.A. and Smyth, R., 2023. High-speed internet access and energy poverty. Energy economics, 127, p.107111. [Published Version] 

• We examine the effects of high-speed internet access on energy poverty in Australia. We find that the expansion of NBN access increases the probability of energy poverty, primarily driven by the decline in social capital resulting from greater internet access.

Hewamalage, H., Ackermann, K. and Bergmeir, C., 2023. Forecast Evaluation for Data Scientists: Common Pitfalls and Best Practices. Data Mining and Knowledge Discovery.  [Published Version] 

• A tutorial style paper on how to choose the evaluation metrics for your forecasting problem. It also emphasis the point on how to evaluate economic and financial time-series.

Ackermann, K., Awaworyi Churchill, S. and Smyth, R., 2023. Broadband internet and cognitive functioning. Economic Record, 99(327), pp.536-563. [Published Version] 

• We examine the effects of high-speed internet access on cognitive functioning in Australia, finding that high-speed internet access causes a decline in crystallised intelligence and that the effects are mediated by social capital and moderated by age and gender. 

Ackermann, K., Churchill, S.A. and Smyth, R., 2021. Mobile phone coverage and violent conflict. Journal of Economic Behavior & Organization, 188, pp.269-287. [Published Version]

• We examine the effects of mobile phone coverage on violent conflicts in Africa. We find that mobile phones expansion increase non-state based conflict, but the effect is mostly driven by the economic inequality that arises of the economic growth due to mobile coverage.

Artificial Intelligence in Medicine

Spencer, L., Fernando, J., Akbaridoust, F., Ackermann, K. and Nosrati, R., 2022. Ensembled Deep Learning for the Classification of Human Sperm Head Morphology. Advanced Intelligent Systems, p.2200111. [Published Version] 

• In vitro fertilisation (IVF) requires the manual selection of a sperm cell. In this work we present a machine learning based system to aid the selection process.

Artificial Intelligence in Public Policy

Ackermann, K., Walsh, J., De Unánue, A., Naveed, H., Navarrete Rivera, A., Lee, S.J., Bennett, J., Defoe, M., Cody, C., Haynes, L. and Ghani, R., 2018, July. Deploying machine learning models for public policy: A framework. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 15-22). [Published Version] 

• We present a framework how to deploy machine learning model as a system to be integrated in the operational workflow in public and private institutions.

Helsby, J., Carton, S., Joseph, K., Mahmud, A.,  Park, Y., Navarrete, A., Ackermann, K., Walsh, J., Haynes, L., Cody, C. and Patterson, M.E., 2018. Early intervention systems: Predicting adverse interactions between police and the public. Criminal justice policy review, 29(2), pp.190-209. [Published Version] 

• We investigate how machine learning can be use to help with stress interventions of police officers compared to a rule based strike system.

Ackermann, K., Blancas Reyes, E., He, S.,  Anderson Keller, T., Van Der Boor, P., Khan, R., ... & González, J. C. (2016, August). Designing policy recommendations to reduce home abandonment in Mexico. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 13-20). [Published Version]

• We build a model for risk assessment for low income households at risk of abandoning their mortgage and consequently their home.

Ackermann, K., & Angus, S. D. (2014). A resource-efficient big data analysis method for the social sciences: the case of global IP activity. International Conference on Computational Science, Procedia Computer Science, 29, 2360-2369.  [Published Version]

• We present a method on how to repurpose a large data-set on internet measurements for insights creation regarding human behaviour.

Media coverage

Teaching

• Data Visualisation and Analytics (Summer Semester 2019, 2020, 2021, 2022, 2023) at Monash University

• Statistical Machine Learning (Semester 2 2019, 2020, 2021, 2022,2023 ) at Monash University

• Harnessing Big Data for Business and Society (Semester 2 2023) at Monash University

• Short Course - Statistical Machine Learning (Summer Semester 2024) at the University of Zürich

Get in touch at klaus.ackermann@monash.edu