Dr Atikur Khan is a Data Scientist who has worked at Qantares, CSIRO and Monash University, Australia. After graduating from Monash University with a PhD in Econometrics and Business Statistics, he started working at CSIRO, Australia. He has developed several statistical methods for forecasting and extracting signals from noisy time series data. His domain of research extended further to statistical learning and statistical disclosure control with special attention to collection, dissemination and analysis of customer data. Some of his research works have been published in top tier journals such as Australian & New Zealand Journal of Statistics, Data & Knowledge Engineering, International Journal of Forecasting, Journal of the American Medical Informatics Association, Journal of Time Series Analysis, and Statistics & Computing.
PhD in Econometrics and Business Statistics: Monash University, Australia, 2013.
MSc in Statistics: National University of Singapore, Singapore, 2008.
- Time series analysis and forecasting: develop new methods and implement those methods by using simulation experiments and real data analysis.
- Statistical disclosure control of business data (statistical databases): collection, dissemination and mining of confidential business data.
- Application of machine learning methods in high frequency and granular data mining, time series analysis and forecasting.
Data Analytic Skills
- Forecasting and time series models (ARIMA, VAR, SSA, GARCH, etc. )
- Machine learning (decision tree, gradient boosting, random forest, support vector machine, etc.)
- Marketing analytics (propensity models, ROI, sentiment analysis, etc.)
- Privacy and confidentiality protection (PIA, data anonymisation, SDC methods, etc.)
- R: used extensively for my research works and able to develop new R package.
- Python: used for my research projects to process and analyse data
- Apache Spark, PySpark, Scala, Kafka, Zookeeper: used for stream processing
- MySQL: used for database management
- Git & GitHub: used for version control