A variety of new tools for data science have been recently added to MATLAB. These include functions for exploratory data analysis and apps for quick exploration of machine learning models and deployment to multiple platforms.
In this session, we explore the fundamentals of data science using MATLAB. We will use an example to address a typical data science problem including data access, pre-processing, machine learning model development, and finally deployment of the model to an application.
Potential topics within this talk include the following:
Accessing and exploring large data sets
Preprocessing and analyzing various types of data including textual and time-stamped data using MATLAB data types: table, timetable, string, categorical, datetime, duration, tall arrays
Working with messy data including outliers, missing, and noisy data, joining tables, synchronizing data by time, and calculating statistics by group
Visualizing various data types: time series plots, heatmaps, wordclouds, geographic plots, boxplots
Training and validating machine learning models using the Classification Learner and Regression Learner apps
Integrating model predictions into a web application running on the cloud
About the Presenter
Hanim is a new addition to the Application Engineering team in TechSource Systems, the sole distributor of MATLAB and Simulink in South East Asia. Previously, she graduated from University of Nottingham with an M.Eng. in Electronic and Communications Engineering and has recently passed her Ph.D. in Electrical and Electronics Engineering. Having been done research projects involving programming works, Hanim has written source codes using MATLAB and utilized its toolboxes for more than 7 years, including her novel and robust evolutionary algorithm, which is an optimization technique of artificial intelligence (AI).
Dr. Hanim Basarudin
Application Engineer, TechSource Systems