Trinh Vu Linh Chi
Cyber-attack Detection for Water Treatment Plant
• Led a team of 3 to complete project within a week as part of an internal Kaggle competition for a school module • Pre-processed of SCADA historical data in R and Python for anomaly detection in stationary multivariate time series data • Employed Tableau in visualisation of past data to study historical trends • Implemented machine learning algorithms for anomaly detection in time series in R and Python, including classification, moving average, Kalman filtering and LSTM with Keras package to model expected behaviour and generate stochastic outlier flags