Dr Chris Dada has added to his considerable range of microbiology and water-quality modelling skills by successfully completing a Postgraduate Certificate programme at Massey University with courses on a range of data analysis and modelling methods. During the courses, Chris extensively applied Python, SAS Enterprise Guide and SAS Enterprise Miner, which are sophisticated programming tools especially suited for big-data analysis and visualisation.
In the Data Mining course, Chris applied SAS Enterprise Miner for classification and regression-tree modelling, market basket and association analysis, neural network modelling, ensembles modelling and text mining.
The Multivariate Big-Data Analysis course covered outlier analysis and detection, data visualization using ordination methods (e.g. principal components analysis, vector overlays and multidimensional scaling), redundancy analysis, canonical correlation analysis, discriminant function analysis, and partial least squares regression.
In Econometrics, Chris applied computer-based specification for estimation and validation of linear and multiplicative regression models, logit models, and time-series models for policy analysis and forecasting.
A key component of all the courses was managing databases and working with real-world datasets, which included data cleaning, wrangling and transformation. Students were required to report their results, answering key policy or industry-relevant questions. For his final project on Multivariate Big-Data Analysis, Chris analysed nearly 40 years (1975–2013) of New Zealand water quality data (750,000 rows by 14 columns of data!) using a range of techniques, to reveal some interesting insights.
Chris’ results suggest that authorities need to look more at expanding programmes aimed at improving or restoring small water bodies or waters flowing through smaller catchments. Chris will talk more about this in future blogs.
Chris received excellent feedback from the lecturers in charge of the data courses, including a request from one lecturer for his permission to use one of Chris’ reports as a template for future students.
Well done Chris!